Guide to the Combined Dataset - Longitudinal · PDF fileField administration: Elaine Iffland, Theresa Patterson, Glenis Naudo, Jill Penn, Maureen Slater CAPI/CASI programming:

NCDS/BCS70 1999-2000 Follow-ups

Guide to the Combined Dataset (June 2001)

Compiled by on behalf of the Joint Centre for Longitudinal Research Team

by

Peter Shepherd

NCDS/BCS70 1999-2000 Follow-ups: Guide to the Combined Dataset (June 2001) – i

JOINT CENTRE FOR LONGITUDINAL RESEARCH The 1999-2000 follow-ups of the National Child Development Study (NCDS) and the 1970 British Cohort Study (BCS70) were carried out under the auspices of the Joint Centre for Longitudinal Research. The JCLR exists to promote and facilitate the widest possible use of longitudinal studies, and the development of expertise in longitudinal research. It is a partnership of three institutions with unparalleled experience in the conduct of longitudinal surveys and the analysis of longitudinal data: • Centre for Longitudinal Studies (CLS), Institute of Education, University of London • International Centre for Health and Society (ICHS), University College Medical School, London • National Centre for Social Research (NatCen) The work of designing and conducting the NCDS and BCS70 1999-2000 Follow-ups was carried out by a joint team drawn from two of the JCLR partners – CLS and the National Centre. The key members of the team are listed below: CLS NCDS/BCS70 Team Research design and implementation:

John Bynner, Peter Shepherd, Elsa Ferri, Kate Smith, Neville Butler

Cohort Studies Administrator: Denise Brown Tracing Unit: Kevin Dodwell, Farnaz Farahmand, (and temporary tracing staff

to numerous to name here) IT Team: Mahmood Sadigh, Mos Mojaddad National Centre NCDS/BCS70 Team Design and implementation of surveys:

Debbie Collins, Kavita Deepchand, Rory Fitzgerald, Jane Perry

Field administration: Elaine Iffland, Theresa Patterson, Glenis Naudo, Jill Penn,

Maureen Slater CAPI/CASI programming: Michael Hart, Kevin Palmer, Steve Kelly The Centre brings together expertise the conduct of longitudinal research and the potential for major scientific advances at the interface of education, medicine and social science.

NCDS/BCS70 1999-2000 Follow-ups: Guide to the Combined Dataset (June 2001) – ii

ACKNOWLEDGEMENTS We wish to acknowledge the support for these two follow-ups of our principal funders: the Economic and Social Research Council; Government Departments and Agencies (Office of National Statistics, Department for Education and Employment, Department of Social security, Department of Health, Scottish Executive, Basic Skills Agency); and the International Centre for Child Studies The work could not have been carried out successfully without the involvement of over a hundred advisors drawn from researchers, policy makers and funders, who we consulted throughout the design of the surveys. The names of NCDS/BCS70 Advisors, and an indication of their contribution may be found in the following document, which also accompanies the data deposit:

The design and conduct of the 1999-2000 surveys of the National Child Development Study and the 1970 British Cohort Study.

NCDS/BCS70 1999-2000 Follow-ups: Guide to the Combined Dataset (June 2001) – iii

PREFACE This document has been prepared to accompany the initial deposit, with the UK Data Archive at the University of Essex, of data from the most recent follow-ups of two continuing, multidisciplinary, national, longitudinal studies – the National Child Development Study (NCDS) and the 1970 British Cohort Study (BCS70). The other elements of the deposit, to which reference will be made throughout this document, are identified below. Users are advised that they will need to consult all elements of the documentation to gain a full understanding of the data.

NCDS/BCS70 Deposit: Elements Title Format NCDS and BCS70 1999-2000 Follow-ups: Initial Cross-sectional Data (June 2001) SPSS The design and conduct of the 1999-2000 surveys of the National Child Development Study and the 1970 British Cohort Study

Word

NCDS/BCS70 1999-2000 Follow-ups: CAPI Documentation Word NCDS/BCS70 1999-2000 Follow-ups: Technical Report Word NCDS/BCS70 1999-2000 Follow-ups: Interactive Data Dictionary for Combined NCDS/BCS70 SPSS dataset (based on the SPSS Data Dictionary)

Idealist for Windows

NCDS/BCS70 1999-2000 Follow-ups: Guide to the Combined Dataset (June 2001) Word NCDS Publications list 2001 Word BCS70 Publications List 2001 Word

NCDS/BCS70 1999-2000 Follow-ups: Guide to the Combined Dataset (June 2001) – iv

CONTENTS Page Joint Centre for Longitudinal Research I Acknowledgements Ii Preface Iii Introduction 1 NCDS/BCS70 1 Advisors 2 Survey instruments/CAPI 2 Content of surveys 3 Fieldwork 6 Data coding and editing 7 Timetable 8 NCDS/BCS70 Data 9 Variable names 9 Variable labels 10 Value labels 11 Missing values 11 Variable order 11 Values 11 Consistency 11 Some useful variables 13 Self-completion 14 Proxy interview 15 Identifiers 16 Reference dates for retrospective data/histories 17 Date of interview 17 CAPI Derived variables 19 Further assessment of data quality 22 Household 24 Housing 25 Relationships 26 Children 30 Family Social Relationships & Support 32 Family Income 33 Employment 34 Lifelong Learning 39 Health 63 Citizenship and Values 82 Self-completion (covering attitudes/especially sensitive topics) 83 Appendices (for details see overleaf) 93

NCDS/BCS70 1999-2000 Follow-ups: Guide to the Combined Dataset (June 2001) – v

Appendices Page Appendix 1 Pregnancy History Variables A2 Appendix 2 Proxy Interview Variables A5 Appendix 3 NCDS/BCS70 Advisors contributing to quality assessment of NCDS/

BCS70 data A10

Appendix 4 Earnings and related variables A11 Appendix 5 Basic Skills and other variables A57 Appendix 6 Problems with reporting of GCSEs A96 Appendix 7 Highest Qualification A138 Appendix 8 Attitude Scales A163

NCDS/BCS70 1999-2000 Follow-ups: Guide to the Combined Dataset (June 2001) – 1

INTRODUCTION This document has been prepared to accompany the initial deposit, with the UK Data Archive at the University of Essex, of data from the most recent follow-ups of two continuing, multidisciplinary, national, longitudinal studies – the National Child Development Study (NCDS) and the 1970 British Cohort Study (BCS70). The follow-ups took place between November 1999 and September 2000, and it is noteworthy that this was the first time that both cohorts had been surveyed at the same time. They were designed and implemented jointly by the Centre for Longitudinal Studies of the Institute of Education, University of London (CLS), and the National Centre for Social Research (NatCen), on behalf of the Joint Centre for Longitudinal Research. The work was mainly funded by the Economic and Social Research Council, but important contributions were also made by a number of government departments, and by the Basic Skills Agency. NCDS and BCS70 The National Child Development Study (NCDS) started life as the Perinatal Mortality Survey and examined the social and obstetric factors associated with stillbirth and infant mortality among over 17,000 babies born in Britain in the week 3-9 March 1958. Since this first study the whole cohort have been surveyed on five other occasions in order to monitor their health, education, social and economic circumstances. These surveys were carried out in 1965 (age 7), 1969 (age 11), 1974 (age 16), 1981 (age 23) and 1991 (age 33). As part of the 1991 survey, a special study was also undertaken of the children of one third of the cohort members, including assessments of the behaviour and cognitive development of approximately 5,000 children. There have also been surveys of sub-samples of the cohort, the recent occurring in 1996 (age 37) when information was collected on the basic skills of a representative sample of 10 per cent of cohort members. The 1970 British Cohort Study (BCS70) was designed along similar lines to the NCDS, surveying over 17,000 babies born in Britain in the week 5-11 April 1970. Since the birth survey there have been four other major data collection exercises in order to monitor their health, education, social and economic circumstances. These were carried out in 1975 (age 5), 1980 (age 10), 1986 (age 16) and 1996 (age 26). As in NCDS, subsamples have been studied at various ages: for example at age 21, paralleling the NCDS survey at age 37, a 10 per cent representative sample was assessed for basic skills difficulties. From their original focus on the circumstances and outcomes of birth, the two cohort studies have broadened in scope to map all aspects of health, education and social development of their subjects as they passed through childhood and adolescence. In latter sweeps, the information collected has covered their transitions into adult life, including leaving full-time education, entering the labour market, setting up independent homes, forming partnerships and becoming parents. The latest rounds of data collection for NCDS and BCS70 took place in 1999/2000 (NCDS cohort member were 41/42 and BCS70 cohort members were 29/30). The main aim of these most recent surveys was to explore the factors central to the formation and maintenance of adult identity in each of the following domains:

• Lifelong learning • Relationships, parenting and housing • Employment and income • Health and health behaviour • Citizenship and values


Further details of this ‘life course’ theoretical framework and its use in the development of question areas for the most recent sweeps of NCDS and BCS70 are contained in the following, which also accompanies the data deposit:


Advisors As with previous NCDS and BCS70 follow-ups, the surveys were designed in collaboration with advisors drawn from researchers, policy makers and funders. Following an initial meeting, held in March 1998, a number of advisory groups were formed, one for each of the proposed major topic areas to be covered by the new surveys – see below.

NCDS/BCS70 Advisory Groups

Child development & education Citizenship and values

Employment and income Family, parenting and housing

Health Lifelong learning

Methodology

After the initial meeting, members of each group exchanged ideas via email and/or meetings and, ultimately provided written advice on the content of the survey instrumentation (June-September 1998). Subsequently, members of the advisory groups were updated on the development of the instrumentation, and the progress of the surveys. Latterly, a number of advisors have been involved in the initial assessment of the quality of the information obtained during the survey. Further details of the advisory groups, and their consultations and advice are contained in the following, which also accompanies the data deposit:


Survey instruments/CAPI Information was gathered from NCDS and BCS70 cohort members by interview and using self-completion questionnaire. A major innovation in the conduct of the new surveys was the adoption of computer-assisted, rather than paper-based, methods of data collection. The use of CAPI (Computer Assisted Personal Interviewing) serves to improve the quality of the data collected, by simplifying the conduct of the interviews with complex filter structures. It also provides for the rapid production of clean data, because of the facility to edit data on entry. In addition, the employment of CAPI simplifies coding of information about occupation and answers to other open-ended questions where responses can be keyed in during the interview. The same method, CASI (Computer Assisted Self-Interviewing), was employed with the self-completion instrument, when the laptop computer was handed over to the cohort members themselves. Following development work - including: pre-piloting qualitative interviews and a pilot based on traditional paper interview schedules and self-completion questionnaire all based on non-cohort member samples; and a CAPI/CASI ‘dress rehearsal’ pilot based on a cohort member sample - the


survey instruments were programmed into a single CAPI/CASI instrument using Blaise 3 which was suitable for both NCDS and BCS70 cohort members. This was possible because over 90 per cent of questions were common to both cohorts. The major differences between the content of the NCDS and BCS70 surveys were: • Reference dates for retrospective questions/histories. These were March 1991 for NCDS, and

April 1986 for BCS70 - although NCDS cohort members who did not take part in the 1991 follow-up (NCDS5), where asked for details of qualifications gained since March 1981 (see below).

• The inclusion of additional questions for NCDS cohort members dealing with:

• Children over the age of 16 years • Children absent from the household, but who were living with cohort member in 1991

A short proxy interview was also included in the Blaise program for use where the cohort member was unable to understand or respond to questions put by the interviewer, or to the self-completion. Questions were put to a family member or carer. Details of the survey instruments and their development are to be found in the following, which also accompanies the data deposit:

NCDS/BCS70 1999-2000 Follow-ups: CAPI Documentation

NCDS/BCS70 1999-2000 Follow-ups: Technical Report Content of surveys As noted above, the survey instrumentation was developed in consultation with those who have been involved with the design and analysis of earlier NCDS and BCS70 surveys, other research advisors and funders; and in accordance with the following principles:

Relevance to the stage of life reached Continuity with previous surveys Comparability across NCDS and BCS70 Compatibility with other surveys (eg: BHPS, the General Household Survey and the (US)

National Longitudinal Survey of Youth) A summary of the topics covered by the surveys is given below.

Survey topics

Household Housing

Relationships Children

Family Social Relationships & Support Family Income Employment

Lifelong Learning Health

Citizenship and Values Self-completion (covering

attitudes/especially sensitive topics)


A more detailed summary of the content of the surveys is also given below, and full details of the content of the survey instruments are to be found in the following, which also accompanies the data deposit:


NCDS/BCS70 Follow-ups 1999/2000 Summary of the contents of the surveys

INTERVIEW Household Grid Household grid Fairly standard grid. Gathers details of sex, age,

relationship to respondent, marital status Ethnicity Based on the new question being developed for the 2001

Census Language spoken in the home Language usually spoken Housing Current address Rooms, heating, housing benefit, tenure Intentions to move Moving, future tenure, why moving Property inheritance Ever inherited house/flat, when Homelessness Experience of, number of times, date of last, why, where

stayed, duration, applied to LA Housing history Works backwards from last address. Seeks details of:

dates, tenure, living arrangements, reasons for moving. Relationships Marital status Legal status, prior cohabitation, dates of

marriage/cohabitation, age, marital status of partner Relationship history Marriage/cohabiting. Works backwards from last

relationship. Seeks details of: dates, sex, age, marital status, marriage, separation/divorce

Children Pregnancy history Works backwards from last pregnancy conceived/

fathered. Seeks details of: outcome, name, sex, birth weight, DOB, problems, smoking in pregnancy, other parent, whereabouts, support and circumstances of absent children

Lone parenthood Periods of ≥1 month. Works forward for up to 4 periods. Seeks details of dates and numbers of children

Infertility Ability to have and plans for (more) children Adopted children Age of child on adoption and nature of adoption Partner’s children from a previous relationship

Number, whether seen or visit

Children over 16 (NCDS only) Name, age, economic, marital and parent status Absent formerly resident children (NCDS only)

Name, age, whereabaouts, economic, marital and parent status; CM contact and payments for children absent from the household, but who were living with CM in 1991

Family activities Things done as a family Demands of parenting Physical/emotional/time demands, worries, closeness of

family Family, Social Relationships & Support Contact with family Parents alive/contact/close/divorce/worries/help. Contact

with siblings/in-laws/grandparents Emotional support Sources/nature of emotional support Family Income Other Income Income from benefits and regular income form other


sources received by respondent and/or partner Financial situation Organisation of household money and money problems Employment Economic activity Fairly standard question Current job Seeks details of: earnings, hours, fringe and

Other benefits, pensions, prospects Other paid work Weekly earnings from odd/casual jobs Currently unemployed How became unemployed and job search Labour market histories Periods in a job or not in a job lasting ≥1 month. Works

backwards. Seeks details of: circumstances and dates. If in job gathers details of job, employer, responsibilities (for SOC and SIC coding)

Partner’s job Age left full-time education, economic status, job, employer, responsibilities (for SOC and SIC coding), earnings

Lifelong Learning Qualifications Age left full-time education and educational and vocational

qualifications held (subjects, grades, dates awarded, where studied)

Current course for qualification Qualification, subject(s), date started, where studied Assessment of current/most recent course

Why taken and expected/experienced benefits

Other courses and training Number, why taken and expected/experienced benefits No formal learning Reasons why no learning Learning overview Useful and enjoyable periods of learning Contact with information technology Use of computers at home and at work Literacy and numeracy Problems with reading, writing and maths, implications and

courses to improve Health General health Self-assessed health and experience of list of conditions,

including age at onset and contact with doctor Long-term health conditions Details of longstanding illnesses, etc (including limiting

impact and age at onset), impact on employment, registered disabled.

Respiratory problems Coughing, phlegm and shortness of breath Mental health Experience of mental health problems, including age at

onset Seeing and hearing Problems with sight/eyes and with hearing Other conditions Details of other health conditions requiring regular medical

supervision Accidents/injuries Works backwards. Details of accidents/ injuries/assaults

(age, why admitted, out/in-patient, type of injury). Nature of any permanent disability resulting from any accident/etc

Hospital admissions Works backwards. Age and why admitted Smoking Smoking habit of respondent and partner Drinking Alcohol consumption in last 7 days, other aspects of

drinking behaviour Diet Frequency of consumption of types of food, vegetarian or

other special diets Exercise Exercise at work and in daily life Height and weight Self-reported height and weight and assessment of weight Citizenship and Values Involvement with organisations, voting behaviour and

intentions, political alignment, trade union membership, religion, newspaper readership, car ownership, values, political activity


SELF-COMPLETION Your views Attitude statements How you get on with your husband, wife or partner

Includes Locke-Wallace

Some more of your views Attitude statements How you feel Malaise Inventory Your skills How good at skill/is skill used at work More of your views Attitude statements How you feel about your life so far GHQ 12 More of your views Attitude statements School exclusion and truancy Number of temporary/permanent suspensions/exclusions;

frequency of truancy Contact with the police and crime Number of times moved on, questioned, warned, taken to

police station, cautioned, found guilty by a court Use of illegal drugs Whether tried number of specific drugs ever/in last 12

months Fieldwork

The main fieldwork for the NCDS and BCS70 surveys began following the first interviewer briefing at the end of October 1999. It was conducted in a series of 6, overlapping waves as shown below.

Dates of Fieldwork Waves

Wave

1st Briefing

Fieldwork

1 29/10/99 29/10-13/12 2 29/11/99 29/11-10/01 3 06/01/00 06/01-14/02 4 31/01/00 31/01-13/03 5 28/02/00 28/02-10/04 6 03/04/00 03/04-25/09

With the exception of the last wave, each wave broadly covered all areas of the country and targeted some 2,500 members of each cohort (5,000 in total). The last wave contained the vast majority of outliers (addresses in the Highlands and islands), as well as movers and others who could not be interviewed in waves 1-5. Each wave began with the mailing of an advance letter advising the cohort members selected for inclusion in the wave that an interviewer would be calling shortly. This was followed by a series of face-to-face interviewer briefings, held in different parts of the country, during which: the background and purpose of the survey was explained; instructions were given on the contact, tracing and other administrative procedures; the survey content was outlined; and the interviewers given an opportunity to conduct a full ‘dummy’ interview (designed to take them through all the main sections of the CAP/CASI instruments, and to highlight areas of questioning and/or issues of definition, where the surveys differ from other surveys on which the interviewer may have worked/be working). During fieldwork, National Centre interviewers administered the CAPI/CASI instruments, after carrying out any necessary tracing to establish the whereabouts of cohort members. The interviewer tracing supplemented the efforts of the small Tracing Team maintained by CLS during the preparations for and conduct of the survey. Between follow-ups efforts are made by CLS, through the mailing of an annual birthday card and other activities, to maintain contact with as many members as possible of both cohorts. But, unfortunately, at any one time an important minority of NCDS/BCS70 cohort members remains untraced, and considerable efforts have been made before and during the surveys to locate as many


as possible of the untraced. This serves not only to maximise response, but also to minimise response bias. Prior to fieldwork, the CLS Tracing Team attempted to obtain a current address for as many cohort members as possible. The work of the team built on experience gained during the NCDS 1981 and 1991 follow-ups, and the BCS70 1996 Postal Follow-up to trace attempt to trace as many cohort members as possible. It made use of a variety of sources of information, as indicated below both before and during fieldwork. During the period of fieldwork, which ended on 25 September, the efforts of the tracing team were supplemented, where necessary, by interviewers who sought to establish the whereabouts of the cohort member, speaking to neighbours and others, and follow-up leads as appropriate. Where the interviewer failed to find the cohort member, information was passed back to the CLS tracing team for further investigation.

Sources of address information during tracing

Annual birthday card mailings Address and contact address information provided by cohort members in the past Other information contained in study records Telephone number databases Postcode databases Electoral register databases National Health Service Central Register records of NHS registration, emigrations and deaths Health Authorities address records Driver and Vehicle Licensing Agency address records Ministry of Defence records Media appeals Interviewer detective work during fieldwork

Further details of the fieldwork and tracing are to be found in the following, which also accompanies the data deposit:

NCDS/BCS70 1999-2000 Follow-ups: Technical Report Data coding and editing Data were returned from the field (via modem) and coding and residual editing undertaken A major advantage of the use of CAPI and CASI is the reduced need for post-fieldwork editing – the majority of checks for validity, range and consistency can be incorporated into the CAPI/CASI program. Inevitably, however, there were checks, which were overlooked, or not initially thought necessary. These checks were incorporated into the DP activities undertaken by the National Centre after the survey. The NCDS/BCS70 interview and self-completion include a number of open-ended questions where the verbatim answers of cohort members are keyed by interviewers, and a rather larger number of questions where precodes are provided for answers but provision is also made to record additional information where then precode ‘other’ is used. Following the start of the surveys, these questions were reviewed by the CLS team in order to determine the priorities for coding, and to identify the appropriate coding frames. Where possible, coding frames that had been employed for earlier NCDS/BCS70 surveys were adopted, although it was usually necessary to include additional codes. In other instances, it was necessary to develop a coding frame from scratch. Coding was undertaken by the National Centre and CLS, with the latter being responsible for coding of health and related


problems using the WHO International Classification of Diseases, and OPCS Classification of Surgical Procedures and Operations. Further details of the editing and coding are to be found in the following, which also accompanies the data deposit:

NCDS/BCS70 1999-2000 Follow-ups: Technical Report Timetable An indication of the overall timetable for the NCDS and BCS70 follow-ups is given below with reference to a number of key events and activities mentioned above. It should be noted that the full funding necessary to carry out both follow-ups was not finally secured until early 2000, and that the main fieldwork period was extended in order to ensure that as many cohort members as possible had an opportunity to participate.

Key event/activity Date Start of survey design and tracing of cohort members January 1998 Consultative conference – advisory groups established March 1998 Written advice received on content from advisory groups June – September 1998 Development of instrumentation begins July 1998 Further consultation with advisors July 1998-July 1999 Qualitative pre-piloting undertaken December 1998-February 1999 First, paper pilot (non-cohort sample) July 1999 Development of CAPI/CASI program July-October 1999 Development of instrumentation completed August 1999 Second, CAPI/CASI pilot - ‘dress rehearsal’ (cohort sample) September 1999 First briefings for main surveys October 1999 Main fieldwork for both follow-ups begins November 1999 Coding frames for open answers agreed, coding begins March 2000 Last briefings for main surveys April 2000 Meeting with advisors to report progress May 2000 Fieldwork ends/tracing of cohort members ends September 2000 Coding and editing completed/Last data transferred to CLS December 2000 Initial assessment of data by CLS and advisors begins January 2001 Deposit of initial cross-sectional data for both follow-ups July 2001


NCDS/BCS70 DATA The main data for the recent NCDS and BCS70 follow-ups is supplied to the UK Data Archive in the form of a single, combined SPSS dataset. This holds 5,188 variables for a total of 22,680 cases – 11,419 NCDS and 11,261 BCS70. As the table below shows, the bulk of these (over 98%) where represent full interviews accompanied by a self-completion. The majority of the remainder are full interviews, where the self-completion was not answered, although there are also a small number of incomplete or partial interviews. Short proxy interviews were undertaken with a family member or carer where the cohort member was unable to understand or respond to questions put by the interviewer or to the self-completion. On a small number of occasions where the cohort member could understand and respond with the aid of an interpreter, an interview was attempted where a family member or carer was able to act as an intermediary.

NCDS/BCS70: Full, partial and proxy interviews

NCDS BCS70 Total Full interview & self completion 11,281 11,116 22,397 Full interview, no self-completion 94 88 182 Partial interview 13 22 35 Proxy interview 31 35 66 Total 11,419* 11,261* 22,680

• 10 NCDS and 7 BCS70 interviews where carried out with the aid of an interpreter (normally another family member)

Source: Crosstabulation of UNOUT x SAMPLE x INTWHO

Variable names - The variable names on the dataset are those automatically allocated by the CAPI program (Blaise 3). Within the Blaise, each question has a variable name (rather than number), made up of a maximum of 8 characters, and this is used to determine the variable name on the dataset. Where the question is repeated (eg: the same question is asked for each birth, relationship, job, qualification, etc reported), Blaise automatically allocates a number suffix (eg: name, name2, name3, name4). Unfortunately, where the variable name in the Blaise program was originally more than 6 characters long, Blaise truncates the name to allow for the suffix. As a result, there is not always a simple match between the Blaise program documentation and the data. A somewhat extreme example of the range of variable names that may be encountered is given in the table below. In the interview, the question: “Who is the other parent of (name of baby)?” was repeated for each child conceived in each pregnancy. In Blaise, and in the CAPI documentation, this question has the name ‘WHOPARB’. In the dataset, variable names are reserved for a maximum of 5 children conceived in each of a maximum of 8 pregnancies. Only for the first baby reported as conceived in the first reported pregnancy does the variable have the name WHOPARB. Blaise allocates modified variable names for each of the 39 other variables which identify the “other parent”. It is important to note that information on pregnancy history was gathered by starting with the most recent pregnancy. A full list of the variables relating to pregnancy history is given in Appendix 1. Details of the CAPII program are to be found in the following, which also accompanies the data deposit:



Variable names for the repeated question: “Who is the other parent of (name of baby)?”

The question is repeated for each child conceived (maximum=5) in each pregnancy (maximum=8) NB: Pregnancy 1 is the most recent pregnancy

Pregnancy Baby Variable name Pregnancy Baby Variable name

1 1 WHOPARB 5 1 WHOPAR33 2 WHOPARB2 2 WHOPAR35 3 WHOPARB3 3 WHOPAR37 4 WHOPARB4 4 WHOPAR39 5 WHOPARB5 5 WHOPAR41

2 1 WHOPARB6 6 1 WHOPAR43 2 WHOPARB7 2 WHOPAR45 - 3 WHOPARB8 3 WHOPAR47 4 WHOPARB9 4 WHOPAR49 5 WHOPAR11 5 WHOPAR51

3 1 WHOPAR13 7 1 WHOPAR53 2 WHOPAR15 2 WHOPAR55 3 WHOPAR17 3 WHOPAR57 4 WHOPAR19 4 WHOPAR59 5 WHOPAR21 5 WHOPAR61

4 1 WHOPAR23 8 1 WHOPAR63 2 WHOPAR25 2 WHOPAR65 3 WHOPAR27 3 WHOPAR67 4 WHOPAR29 4 WHOPAR69 5 WHOPAR31 5 WHOPAR71

Variable labels – The variable labels included on the dataset were also initially derived from the CAPI program. In exporting the SPSS dataset from Blaise, labels based on the wording of questions were automatically allocated. Subsequently, these have been individually reviewed and, where necessary, modified in an effort to ensure that labels are comprehensible and accurate. Again, particular problems occurred where a question was repeated (eg: the same question is asked for each birth, relationship, job, qualification, etc reported). When initially created, the Blaise-generated dataset had identical labels for each repeat of the question. In revising these labels, efforts have been made to indicate which variables relate to which birth, relationship, job, qualification, etc. Again, an example based on the repeated question: “Who is the other parent of (name of baby)?” is given below. Within the label, the “(P1)”, “(P2)”, etc identify the first reported pregnancy, second reported pregnancy, etc.; and the “baby1”, “baby2”, etc identify the first, second, etc baby reported as conceived. It is important to note that information on pregnancy history was gathered by starting with the most recent pregnancy. Similar conventions are used for the other histories within the dataset.

Variable labels for repeated question: “Who is the other parent of (name of baby)?”

NB: Pregnancy 1 is the most recent pregnancy

Pregnancy Baby Variable name Label 1 1 WHOPARB (P1) Who is baby1s other parent 2 1 WHOPARB6 (P2) Who is baby1s other parent 3 1 WHOPAR13 (P3) Who is baby1s other parent 4 1 WHOPAR23 (P4) Who is baby1s other parent 5 1 WHOPAR33 (P5) Who is baby1s other parent 6 1 WHOPAR43 (P6) Who is baby1s other parent 7 1 WHOPAR53 (P7) Who is baby1s other parent 8 1 WHOPAR63 (P8) Who is baby1s other parent


Value labels – The value labels are also similarly derived from the Blaise program and have similarly been reviewed and, where necessary, modified in an effort to ensure that labels are comprehensible and accurate. Missing values – In general, the use of CAPI/CASI has meant that missing data is less common than in earlier NCDS/BCS70 surveys. Missing values are not identified as such within the initial dataset (“declared missing” within SPSS), nor are they labelled. However, “refused”, “don’t know”, “not answered” and “not applicable” have been given consistent values and should be readily distinguishable:

Missing values (unless otherwise labelled)

7, 97, 997, 9997, 99997, 999997 = Refused 8, 98, 998, 9998, 99998, 999998 = Don’t know 9, 99, 999, 9999, 99999, 999999 = Not answered . (sysmis) = Not applicable

Variable order – The order in which variables appear in the dataset will broadly follow the order of sections, and of questions within sections of the survey instruments. However, the order is determined by the structure of the Blaise program, which does not necessarily hold each question in the order in which they are put to the respondent. As the example given in the table below illustrates, the sequence of variables in the dataset relating to drug use does not follow the same order as the questions on drug use in the self-completion (and as shown in: National Centre for Social Research NCDS/BCS70 Team (2000) NCDS/BCS70 CAPI Documentation which also accompanies the data deposit). This change in order is typically, but not exclusively associated with question sequences which are repeated to produce grid-like data structures (eg: birth, relationship, job, qualification histories, etc). Values – As this was a CAPI/CASI survey, the values should be within the specified range for each variable. Consistency – Again, the use of CAPI/CASI should ensure that all filters have been correctly followed. Further details of the content of the data set can be found by generating an SPSS ‘data dictionary’. An interactive version of this which facilitates key word/phrase searches of the content of the dataset also accompanies the data deposit:

NCDS/BCS70 1999-2000 Follow-ups: Interactive Data Dictionary for Combined NCDS/BCS70 SPSS dataset (based on the SPSS Data Dictionary)

EXCLUDED VARIABLES

A number of variables have been removed from the dataset originally derived from the

CAPI/CASI program in order to ensure that the anonymity of cohort members is preserved.


Self-completion (CASI) and dataset order of variables relating to drug use

Self-completion (CASI) order Dataset order

Variable Label Variable Sequentialposition Label

CANNABIS (SC) Have you ever tried cannabis? DRUG 6360 (SC) Name of 1st other drug not mentioned ECSACY (SC) Have you ever tried ecstasy? DRUG12 6365 (SC) Taken 1st other drug in the last 12 mths? AMPHET (SC) Have you ever tried amphetamines? MORE31 6366 (SC) Taken more than 1 other drug? LSD (SC) Have you ever tried LSD? DRUG2 6367 (SC) Name of 2nd other drug not mentioned POPPER (SC) Have you ever tried amyl nitrate? DRUG13 6372 (SC) Taken 2nd other drug in the last 12 mths? MAGMUSH (SC) Have you ever tried magic mushrooms? MORE32 6373 (SC) Taken more than 2 other drugs? COCAINE (SC) Have you ever tried cocaine? DRUG3 6374 (SC) Name of 3rd other drug not mentioned TEMAZ (SC) Have you ever tried temazepan? DRUG14 6379 (SC) Taken 3rd other in the last 12 mths? SEMERON (SC) Have you ever tried semeron? MORE33 6380 (SC) Taken more than 3 other drugs? KETAMINE (SC) Have you ever tried ketamine? DRUG4 6381 (SC) Name of 4th other drug CRACK (SC) Have you ever tried crack? DRUG15 6386 (SC) Taken 4th other drug in the last 12 mths? HEROIN (SC) Have you ever tried heroin? MORE34 6387 (SC) Taken more than 4 other drugs? METHAD (SC) Have you ever tried methadone? DRUG5 6388 (SC) Name of 5th other drug OTHDRUG (SC) Have you tried any other illegal drugs? DRUG16 6393 (SC) Taken 5th other drug in the last 12 mths? DRUG (SC) Name of 1st other drug not mentioned MORE35 6394 (SC) Taken more than 5 other drugs? DRUG12 (SC) Taken 1st other drug in the last 12 mths? DRUG6 6395 (SC) Name of 6th other drug MORE31 (SC) Taken more than 1 other drug? DRUG17 6400 (SC) Taken 6th other drug in the last 12 mths? DRUG2 (SC) Name of 2nd other drug not mentioned MORE36 6401 (SC) Taken more than 6 other drugs? DRUG13 (SC) Taken 2nd other drug in the last 12 mths? [Other CASI variables here] MORE32 (SC) Taken more than 2 other drugs? CANNABIS 6526 (SC) Have you ever tried cannabis? DRUG3 (SC) Name of 3rd other drug not mentioned ECSACY 6527 (SC) Have you ever tried ecstasy? DRUG14 (SC) Taken 3rd other in the last 12 mths? AMPHET 6528 (SC) Have you ever tried amphetamines? MORE33 (SC) Taken more than 3 other drugs? LSD 6529 (SC) Have you ever tried LSD? DRUG4 (SC) Name of 4th other drug POPPER 6530 (SC) Have you ever tried amyl nitrate? DRUG15 (SC) Taken 4th other drug in the last 12 mths? MAGMUSH 6531 (SC) Have you ever tried magic mushrooms? MORE34 (SC) Taken more than 4 other drugs? COCAINE 6532 (SC) Have you ever tried cocaine? DRUG5 (SC) Name of 5th other drug TEMAZ 6533 (SC) Have you ever tried temazepan? DRUG16 (SC) Taken 5th other drug in the last 12 mths? SEMERON 6534 (SC) Have you ever tried semeron? MORE35 (SC) Taken more than 5 other drugs? KETAMINE 6535 (SC) Have you ever tried ketamine? DRUG6 (SC) Name of 6th other drug CRACK 6536 (SC) Have you ever tried crack? DRUG17 (SC) Taken 6th other drug in the last 12 mths? HEROIN 6537 (SC) Have you ever tried heroin? MORE36 (SC) Taken more than 6 other drugs? METHAD 6538 (SC) Have you ever tried methadone? OTHDRUG 6539 (SC) Have you tried any other illegal drugs?


Some useful variables Variables included on the initial dataset, which may be of particular value to users are identified below.

Some useful variables

Information Variables Survey elements completed UNOUT, INTWHO NCDS or BCS70 respondents SAMPLE Self-completion UNOUT, CASIINT (see below) Proxy interviews UNOUT, IFPROXY (see below) Identifiers NSERIAL/BSERIAL (see below) Sex DMSEX, CMSEX Ethic group ETHNIC Date of interview INTDATE/DATEINT (see below) Derived variables within CAPI See below

Additional information about the self-completion, proxy interviews, identifiers and derived variables used within CAPI is given below, along with additional guidance for users.


Self-completion The self-completion (CASI) was administered at the end of the interview. The interviewer handed the laptop computer used for the interview to the cohort member and explained how they should complete the questionnaire. Where the cohort member was unable or reluctant to use the laptop, the interviewer assisted, and if necessary administered the self-completion as an interview. The variables which hold the data for the self-completion are identifiable through their labels – all begin with the endorsement “(SC)”. These variables are identified below.

Variables which hold data for the Self-completion (CASI) (ordered left to right)

DRUG DRUG12 MORE31 DRUG2 DRUG13 MORE32 DRUG3 DRUG14 MORE33 DRUG4 DRUG15 MORE34 DRUG5 DRUG16 MORE35 DRUG6 DRUG17 MORE36 LR1 AR1 E1 A1 C1 PC1 L1 MOR1 LR2 WM1 IT1 WE1 MOR2 HAPPYREL OUTTOG OUTALONE POUTALONSAMEPART WISHP1 WISHP2 WHOCOOKS WHOSHOPS WHOCLEAN WHOWASH WHODIY WHOCASH WHOTENDS WHOTEACH WHOCARES A2 MOR3 LR3 AR2 C2 L2 PC2 LR4 E2 A3 WM2 AR3 GHQ1 GHQ2 GHQ3 GHQ4 GHQ5 GHQ6 GHQ7 GHQ8 GHQ9 GHQ10 GHQ11 GHQ12 SKILL1A SKILL1B SKILL2A SKILL2B SKILL3A SKILL3B SKILL4A SKILL4B SKILL5A SKILL5B SKILL6A SKILL6B SKILL7A SKILL7B SKILL8A SKILL8B SKILL9A SKILL9B MAL01 MAL02 MAL03 MAL04 MAL05 MAL06 MAL07 MAL08 MAL09 MAL10 MAL11 MAL12 MAL13 MAL14 MAL15 MAL16 MAL17 MAL18 MAL19 MAL20 MAL21 MAL22 MAL23 MAL24 LR5 AR4 IT2 MOR4 A4 E3 WE2 LR6 WM4 L3 C3 A5 SUSPSCH NUMSUSPS EXCLSCH NUMEXCLS TRUANT POLICE1 POL1NUM POLICE2 POL2NUM POLICE3 POL3NUM POLICE4 POL4NUM POLICE5 POL5NUM COURT COURTNUM CANNABIS ECSACY AMPHET LSD POPPER MAGMUSH COCAINE TEMAZ SEMERON KETAMINE CRACK HEROIN METHAD OTHDRUG WM3 LR7 MOR5 PC3 L4 AR5 MOR6 A6 IT4 WE3 WM5 C4 IT5 EFFICAC1 EFFICAC2 EFFICAC3 LIFESAT1 LIFESAT2 THANK

Further details of the self-completion are to be found in the following, which also accompanies the data deposit:

NCDS/BCS70 1999-2000 Follow-ups: Technical Report


Proxy interview As noted above, where the cohort member was unable to understand or respond to questions put by the interviewer or to the self-completion, short proxy interviews were undertaken with a family member or carer. The variables which hold the data for the proxy interview are identifiable through their labels – all begin with the endorsement “(Proxy)”. These variables are identified below.

Variables which hold data for the Proxy Interview

(ordered left to right) PROXTHK PROXYWHO PROXYAC1 PROXYAC2 PROXYAC3 PROXYAC4 LSIIMWK LSIAGE LSILIM MORE11 LSIIMWK2 LSIAGE2 LSILIM2 MORE12 LSIIMWK3 LSIAGE3 LSILIM3 MORE13 LSIIMWK4 LSIAGE4 LSILIM4 MORE14 LSIIMWK5 LSIAGE5 LSILIM5 MORE15 LSIIMWK6 LSIAGE6 LSILIM6 MORE16 LSIIMWK7 LSIAGE7 LSILIM7 MORE17 LSIIMWK8 LSIAGE8 LSILIM8 MORE18 LSIIMWK9 LSIAGE9 LSILIM9 MORE19 LSIIMW10 LSIAGE10 LSILIM10 MORE20 HOAGE HOSUP1 HOSUP2 HOSUP3 HOMORE HOAGE2 HOSUP4 HOSUP5 HOSUP6 HOMORE2 HOAGE3 HOSUP7 HOSUP8 HOSUP9 HOMORE3 HOAGE4 HOSUP10 HOSUP11 HOSUP12 HOMORE4 HOAGE5 HOSUP13 HOSUP14 HOSUP15 HOMORE5 HOAGE6 HOSUP16 HOSUP17 HOSUP18 HOMORE6 HOAGE7 HOSUP19 HOSUP20 HOSUP21 HOMORE7 HOAGE8 HOSUP22 HOSUP23 HOSUP24 HOMORE8 HOAGE9 HOSUP25 HOSUP26 HOSUP27 HOMORE9 HOAGE10 HOSUP28 HOSUP29 HOSUP30 HOMORE10 YEARIN YEARM TENURE MARSTAT PREGPROX PROXYTYP PROXYECO CJSUP CJEMPS AGELFTED ACTAGEL FURTHED LFTMORED EDQUALS EDQTYP01 EDQTYP03 EDQTYP04 EDQTYP05 EDQTYP06 EDQTYP07 EDQTYP08 EDQTYP09 EDQTYP10 EDQTYP11 EDQTYP12 EDQTYP13 VOCQUALS VOCTYP01 VOCTYP02 VOCTYP03 VOCTYP04 VOCTYP05 VOCTYP06 VOCTYP07 VOCTYP08 VOCTYP09 VOCTYP10 VOCTYP11 LSIANY LSIREG HEIGHT HTMETRES HTCMS HTFEET HTINCHES WEIGHT WTKILOS WTSTONES WTPOUNDS IFPROXY NMPRX NMPRX2 NMPRX3 NMPRX4 NMPRX5 NMPRX6 NMPRX7 NMPRX8 NMPRX9 NMPRX10 SEXPRX SEXPRX2 SEXPRX3 SEXPRX4 SEXPRX5 SEXPRX6 SEXPRX7 SEXPRX8 SEXPRX9 SEXPRX10 AGEPRX AGEPRX2 AGEPRX3 AGEPRX4 AGEPRX5 AGEPRX6 AGEPRX7 AGEPRX8 AGEPRX9 AGEPRX10 MSPRX MSPRX2 MSPRX3 MSPRX4 MSPRX5 MSPRX6 MSPRX7 MSPRX8 MSPRX9 MSPRX10 RELPRX RELPRX2 RELPRX3 RELPRX4 RELPRX5 RELPRX6 RELPRX7 RELPRX8 RELPRX9 RELPRX10 CMSEXPX BDAT1PX BDAT2PX CMNAMPRX RENAMPRX RESIDPRX NORMPRX Further details of the proxy interview are to be found in Appendix 2, and the following, which also accompanies the data deposit:

NCDS/BCS70 1999-2000 Follow-ups: Technical Report


Identifiers NCDS and BCS70 have unique identifiers which appear on the datasets already lodged with the UK Data Archive. These identifiers are also to be found on the new dataset. • NCDS – The NCDS identifier is given by the variable NSERIAL. This may be used to match the

new data with existing datasets where the same unique identifier will have the variable name SERIAL.

• BCS70 – Two BCS70 identifiers appear on the dataset:

• BSERIAL – This should be used to match the new data with existing BCS70 datasets which use the identifier SERIAL.

• KEY – This should be used to match the new data with existing BCS70 datasets which use

the identifier KEY. It may also be used to match with existing datasets in which the unique identifier is given by a combination of ‘CHES number’ and ‘Twincode’. The former will normally be given the variable name: CHESNO; whilst the latter may variously appear as: TC, CTC, TC2, TC10, etc. To facilitate the match it is important to know that:

KEY = (CHESNO x 10) + TC

CHESNO=trunc(KEY/10)

TC=mod(KEY,CHESNO)

A NOTE OF CAUTION

The initial dataset may be matched with data from earlier NCDS surveys using the unique identifiers included. However, it is important to note that, to date, efforts have concentrated on an internal (cross-sectional) review of the quality of the data, and although longitudinal linkage has been made, there have been no efforts to validate the link through longitudinal editing. Users merging new and old data are strongly advised to carry out their own checks on the validity of the longitudinal link. They should report the details of any problems encountered to the User Support Group via email ([email protected]).


Reference dates for retrospective data/histories Although NCDS and BCS70 are prospective longitudinal studies, the gap between follow-ups has ensured that each includes a number of retrospective questions, which focus on experience since the previous follow-up. The recent follow-ups were no exception; retrospective information was gathered on pregnancies, relationships, jobs, qualifications and health problems since the last major follow-ups as follows: • NCDS: the reference date was March 1991, representing the time of the last major follow-up in

1991 (NCDS5) when the members of the cohort were aged 33 years.

This means that the retrospective histories only cover ages 33 to 41 or 42 (depending on date of interview). Similar histories for ages 16 to 33 are to be found in the NCDS4 and NCDS5 datasets. There is one exception. For those not interviewed at NCDS5, information about qualifications was gathered with reference to March 1981, representing the time of the 1981 follow-up (NCDS4) when the cohort were aged 23 years. In order to compile a complete history from age 16 it will be necessary to link data from the NCDS5 and NCDS6 surveys (and for some it will also be necessary to link data from the NCDS4 survey).

• BCS70: the reference date was April 1986, representing the time of the major follow-up in 1986

when the members of the cohort were aged 16 years.

This means that the retrospective data gathered during the recent follow-up provides histories covering the ages 16-29 or 30 (depending on date of interview). The BCS70 1996 Postal Follow-up was not used to define the reference date because it included very few retrospective questions, and because response to the survey was limited by the need to plan and implement the survey in a limited time interval. This was a consequence of the nature of the funding available.

Date of interview As note above, fieldwork for the NCDS/BCS70 follow-ups took place between November 1999 and September 2000. Each interview included on the dataset was date-stamped by the laptop used by the interviewer to administer the CAPI/CASI instruments. Although, interviewers were asked at the start of the Blaise program to correct any erroneous date or time, subsequent checking revealed that a small number of interview dates were clearly wrong. Fortunately, it has been possible to correct the majority of these with reference to other survey records. However, the dates for 9 interviews remain to be resolved. These are identified below, together with details of the fieldwork wave in which it the address was issued to interviewers. It will be seen that all but one interview was carried out as part of wave 3 which began in January 2000. Whilst it is tempting to make an educated guess about the correct date of interview, experience suggests that it is better to await confirmation. Updates for these cases will be provided as soon as they are available.


Interview dates: Unresolved errors

NSERIAL/

BSERIAL INDATE DATEINT WaveWave start (approx)

NCDS 422081J 15061999 15061999 3 January 2000

518148Y 20122000 20122000 1 November 1999 860002E 09071999 09071999 3 January 2000 860006P 03071999 03071999 3 January 2000 860008T 28061999 28061999 3 January 2000

BCS70 01467071 27061999 27061999 3 January 2000 01610017 03101999 03101999 3 January 2000 03022037 12071999 12071999 3 January 2000 04012039 21061999 21061999 3 January 2000

Queries about these problems should be directed to the User Support Group ([email protected])


CAPI Derived variables Routing respondents through the interview required the derivation of a number of summary variables within CAPI program. As the table below shows, these provide information in a number of areas, about the household, relationships, children, housing, economic activity and qualifications, which may be of value to users.

Some useful CAPI Derived variables

Variable Label HHSIZE Total Number of people in hhld (Derived) AGEP Age of parner [derived] SEXP Sex of partner [derived] NUMADCH No: of adopted kids in hhld [derived] ANYCHD Whether any kids in hhld [derived] CHD16F Any kids aged 0-16yrs in hhld [derived] CHD13F Any kids aged 0-13yrs in hhld [derived] CHDAGE3 Any kids aged 3+yrs in hhld [derived] CHDAGE4 Any kids aged 4-15yrs in hhld [derived] CHD5_16 Any kids aged 5-16 yrs in hhld [derived] CHD16 Any kids 16+yrs in hhld [derived] CHD0_6 (CM NCDS) No: of own kids in hhld aged 0-6yrs [derived] OWNCHILD Whether CM has any of their own kids in hhld [derived] FATHIN Whether CMs natural father lives in hhld [derived] MOTHIN Whether CMs natural mother lives in hhld [derived] NSPOUSE Whether CMs spouse in hhld [derived] NPART Whether CMs partner in hhld [derived] NUMROOMS Total no. rooms (derived) MOVOUT-MOVOUT25 (A1-A25) Date moved out (derived) MOVIN2-MOVIN25 (A2-A25) Date moved in (derived) LIVIN-LIVIN9 (EP1-EP9) Date relationship started (derived) LIVOUT-LIVOUT9 (EP1-EP9) Date relationship ended (derived) NUMP Total number of babies in ref period [derived] LONESTRT-LONESTRT3 Date 1st-3rd period lp started (derived) LONEEND-LONEEND7 Date 1st-3rd period lp ended (derived) ABCHAGE-ABCHANGE8 Age of eldest-8th eldest child living elsewhere (derived) STRTJOB-SRTJO10 PrevAct1-10: Date started activity (derived) CSTRTJOB Date started current activity (derived) NUMGCSE Total no. of GCSE quals CM has (derived) NUMOLVL Total no. of GCE O Levels CM has [derived] NUMCSE Tot no. of CSEs:EDCSE1+EDCSE2 [derived] NUMASLVL Total no. of A/S levels CM has (derived) NUMGCSAS Total no. of GCE A level/S level quals CM has (derived)

All CAPI derived variables include the word ‘derived’ within the label. A full list of CAPI derived variables is given below. More details of the CAPI dereived variables are to be found in the following, which also accompanies the data deposit:



CAPI Derived Variables

HHSIZE Total Number of people in hhld (Derived) MOVIN3 (A3) Date moved in (derived) BIRTHDAY CM day of birthday (derived) MOVOUT3 (A3) Date moved out (derived) AGEP Age of parner [derived] MOVIN4 (A4) Date moved in (derived) SEXP Sex of partner [derived] MOVOUT4 (A4) Date moved out (derived) NUMADCH No: of adopted kids in hhld [derived] MOVIN5 (A5) Date moved in (derived) ANYCHD Whether any kids in hhld [derived] MOVOUT5 (A5) Date moved out (derived) CHD16F Any kids aged 0-16yrs in hhld [derived] MOVIN6 (A6) Date moved in (derived) CHD13F Any kids aged 0-13yrs in hhld [derived] MOVOUT6 (A6) Date moved out (derived) CHDAGE3 Any kids aged 3+yrs in hhld [derived] MOVIN7 (A7) Date moved in (derived) CHDAGE4 Any kids aged 4-15yrs in hhld [derived] MOVOUT7 (A7) Date moved out (derived) CHD5_16 Any kids aged 5-16 yrs in hhld [derived] MOVIN8 (A8) Date moved in (derived) CHD16 Any kids 16+yrs in hhld [derived] MOVOUT8 (A8) Date moved out (derived) CHD16N01 Child reference number [derived] MOVIN9 (A9) Date moved in (derived) CHD16N02 Child reference number [derived] MOVOUT9 (A9) Date moved out (derived) CHD16N03 Child reference number [derived] MOVIN10 (A10) Date moved in (derived) CHD16N04 Child reference number [derived] MOVOUT10 (A10) Date moved out (derived) CHD16N05 Child reference number [derived] MOVIN11 (A11) Date moved in (derived) CHD16N06 Child reference number [derived] MOVOUT11 (A11) Date moved out (derived) CHD16N07 Child reference number [derived] MOVIN12 (A12) Date moved in (derived) CHD16N08 Child reference number [derived] MOVOUT12 (A12) Date moved out (derived) CHD16N09 Child reference number [derived] MOVIN13 (A13) Date moved in (derived) CHD16N10 Child reference number [derived] MOVOUT13 (A13) Date moved out (derived) ADPTNU01 Adopted child reference number [derived] MOVIN14 (A14) Date moved in (derived) ADPTNU02 Adopted child reference number [derived] MOVOUT14 (A14) Date moved out (derived) ADPTNU03 Adopted child reference number [derived] MOVIN15 (A15) Date moved in (derived) ADPTNU04 Adopted child reference number [derived] MOVOUT15 (A15) Date moved out (derived) ADPTNU05 Adopted child reference number [derived] MOVIN16 (A16) Date moved in (derived) ADPTNU06 Adopted child reference number [derived] MOVOUT16 (A16) Date moved out (derived) ADPTNU07 Adopted child reference number [derived] MOVIN17 (A17) Date moved in (derived) ADPTNU08 Adopted child reference number [derived] MOVOUT17 (A17) Date moved out (derived) ADPTNU09 Adopted child reference number [derived] MOVIN18 (A18) Date moved in (derived) ADPTNU10 Adopted child reference number [derived] MOVOUT18 (A18) Date moved out (derived) CHD0_6 (CM NCDS) No: of own kids in hhld aged 0-6yrs [derived] MOVIN19 (A19) Date moved in (derived) OWNCHILD Whether CM has any of their own kids in hhld [derived] MOVOUT19 (A19) Date moved out (derived) FATHIN Whether CMs natural father lives in hhld [derived] MOVIN20 (A20) Date moved in (derived) MOTHIN Whether CMs natural mother lives in hhld [derived] MOVOUT20 (A20) Date moved out (derived) NSPOUSE Whether CMs spouse in hhld [derived] MOVIN21 (A21) Date moved in (derived) NPART Whether CMs partner in hhld [derived] MOVOUT21 (A21) Date moved out (derived) LINENO Line no. of person from hhld grid [derived] MOVIN22 (A22) Date moved in (derived) SEX CM gender [derived] MOVOUT22 (A22) Date moved out (derived) AGE CM age last birthday [derived] MOVIN23 (A23) Date moved in (derived) HSIZE No: of people in hhld [derived] MOVOUT23 (A23) Date moved out (derived) YEARNOW Year interview took place (derived) MOVIN24 (A24) Date moved in (derived) NUMROOMS Total no. rooms (derived) MOVOUT24 (A24) Date moved out (derived) MOVOUT (A1) Date moved out (derived) MOVIN25 (A25) Date moved in (derived) MOVIN2 (A2) Date moved in (derived) MOVOUT25 (A25) Date moved out (derived) MOVOUT2 (A2) Date moved out (derived) LIVIN (EP1) Date relationship started (derived)


CAPI Derived Variables (continued)

LIVOUT (EP1) Date relationship ended (derived) LINEGRI9 9th adopted child listed in hhld grid (derived) LIVIN2 (EP2) Date relationship started (derived) LINEGR10 1st child 16+ in hhld grid (derived) LIVOUT2 (EP2) Date relationship ended (derived) LINEGR11 2nd child 16+ in hhld grid (derived) LIVIN3 (EP3) Date relationship started (derived) LINEGR12 3rd child 16+ in hhld grid (derived) LIVOUT3 (EP3) Date relationship ended (derived) LINEGR13 4th child 16+ in hhld grid (derived) LIVIN4 (EP4) Date relationship started (derived) LINEGR14 5th child 16+ in hhld grid (derived) LIVOUT4 (EP4) Date relationship ended (derived) LINEGR15 6th child 16+ in hhld grid (derived) LIVIN5 (EP5) Date relationship started (derived) LINEGR16 7th child 16+ in hhld grid (derived) LIVOUT5 (EP5) Date relationship ended (derived) ABCHAGE Age of eldest child living elsewhere (derived) LIVIN6 (EP6) Date relationship started (derived) ABCHAGE2 Age of 2nd eldest child living elsewhere (derived) LIVOUT6 (EP6) Date relationship ended (derived) ABCHAGE3 Age of 3rd eldest child living elsewhere (derived) LIVIN7 (EP7) Date relationship started (derived) ABCHAGE4 Age of 4th eldest child living elsewhere (derived) LIVOUT7 (EP7) Date relationship ended (derived) ABCHAGE5 Age of 5th eldest child living elsewhere (derived) LIVIN8 (EP8) Date relationship started (derived) ABCHAGE6 Age of 6th eldest child living elsewhere (derived) LIVOUT8 (EP8) Date relationship ended (derived) ABCHAGE7 Age of 7th eldest child living elsewhere (derived) LIVIN9 (EP9) Date relationship started (derived) ABCHAGE8 Age of 8th eldest child living elsewhere (derived) LIVOUT9 (EP9) Date relationship ended (derived) STRTJOB PrevAct1: Date started activity (derived) NUMP Total number of babies in ref period [derived] STRTJOB2 PrevAct2: Date started activity (derived) LONESTRT Date 1st period lp started (derived) STRTJOB3 PrevAct3: Date started activity (derived) LONEEND Date 1st period lp ended (derived) STRTJOB4 PrevAct4: Date started activity (derived) LONESTR2 Date 2nd period of lone parenthood started (derived) STRTJOB5 PrevAct5: Date started activity (derived) LONEEND4 Date 2nd period of lone parenthood ended (derived) STRTJOB6 PrevAct6: Date started activity (derived) LONESTR3 Date 3rd period of lone parenthood started (derived) STRTJOB7 PrevAct7: Date started activity (derived) LONEEND7 Date 3rd period of lone parenthood ended (derived) STRTJOB8 PrevAct8: Date started activity (derived) LINEGRID 1st adopted child listed in hhld grid (derived) STRTJOB9 PrevAct9: Date started activity (derived) LINEGRI2 2nd adopted child listed in hhld grid (derived) STRTJO10 PrevAct10: Date started activity (derived) LINEGRI3 3rd adopted child listed in hhld grid (derived) CSTRTJOB Date started current activity (derived) LINEGRI4 4th adopted child listed in hhld grid (derived) NUMGCSE Total no. of GCSE quals CM has (derived) LINEGRI5 5th adopted child listed in hhld grid (derived) NUMOLVL Total no. of GCE O Levels CM has [derived] LINEGRI6 6th adopted child listed in hhld grid (derived) NUMCSE Tot no. of CSEs:EDCSE1+EDCSE2 [derived] LINEGRI7 7th adopted child listed in hhld grid (derived) NUMASLVL Total no. of A/S levels CM has (derived) LINEGRI8 8th adopted child listed in hhld grid (derived) NUMGCSAS Total no. of GCE A level/S level quals CM has (derived)


FURTHER ASSESSMENT OF DATA QUALITY In addition to the checks that were built into the CAPI code, or undertaken immediately after fieldwork, members of the CLS NCDS/BCS70 team have carried out further checks. In this task they have been greatly assisted by a number of researchers who are members of the NCDS/BCS70 Advisory Groups and/or their associates and, as such, are experienced in the use of data from previous NCDS/BCS70 follow-ups. It can take some time to become familiar with a large and complex dataset, such as this, and it is important to stress that the checking continues. However, in assessing the quality of the new data, the CLS team and advisors were asked to concentrate on the aspects listed below.

Guide for Quality Assessment

• Variable labels – Check that these are present for all variables, comprehensible and accurate. • Value labels – Check that these are present where appropriate, comprehensive, comprehensible

and accurate. • Values – Report all variables for which the values appear unusual/wrong in any way. • Consistency – Report all instances of apparent inconsistency, eg:

• Where the responses to a primary (filter) question (eg: “Do you have any of the qualifications on this card?”) and supplementary (filtered) question (eg: “IF YES, Which?”) are not consistent.

• Where the respondent’s circumstances (eg: marital status, economic status) are not consistent throughout the dataset.

• Missing values – Report all instances where:

• There are many missing cases. • Missing values are present but not declared or labelled.

• Variable order – Report all instances where confusion occurs because variables appear out of

sequence. • Other problems – Report any and all other problems encountered in using the dataset. • Derived variables – Provide details of any derived variables developed which may of be of value

to other users and which may be considered for deposit with the UK Data Archive.


Some of the information presented above has drawn on lessons learned as a result of this quality assessment. Additional, and more detailed information and guidance is given below, organised by survey topic. A list of the advisors and their associates who contributed to the assessment of the quality of the data, and whose work is drawn on below is given in Appendix 3. The contribution of individual advisors is acknowledged below Wherever appropriate, the detailed documentation provided by advisors is included in Appendices 4-8.

Survey topics

Household Housing

Relationships Children

Family Social Relationships & Support Family Income Employment

Lifelong Learning Health

Citizenship and Values Self-completion (covering

attitudes/especially sensitive topics)


HOUSEHOLD Household Grid Household grid Fairly standard grid. Gathers details of sex, age,

relationship to respondent, marital status Ethnicity Based on the new question being developed for the 2001

Census Language spoken in the home Language usually spoken Review of these data by the CLS NCDS/BCS70 Team and by advisors has, to date, revealed no particular problems, nor have any derived variables been developed. (A number of CAPI dedrived variables are available – see above). However, it should be noted that the household grid includes information about the current de-facto marital status for all persons in the household. For a small number of cohort members, this information has been shown to be inconsistent with information about their legal marital status recorded later in the interview. Further details are reported below under ‘Relationships’. The data will continue to be investigated and, if necessary, detailed updates will be made available, and the dataset will be updated.

Queries about problems should be directed to the User Support Group ([email protected])


HOUSING Housing Current address Rooms, heating, housing benefit, tenure Intentions to move Moving, future tenure, why moving Property inheritance Ever inherited house/flat, when Homelessness Experience of, number of times, date of last, why, where

stayed, duration, applied to LA Housing history Works backwards from last address. Seeks details of:

dates, tenure, living arrangements, reasons for moving. Review of these data by the CLS NCDS/BCS70 Team and by advisors has, to date, revealed no particular problems, nor have any derived variables been developed. The data will continue to be investigated and, if necessary, detailed updates will be made available, and the dataset will be updated.



RELATIONSHIPS Relationships Marital status Legal status, prior cohabitation, dates of

marriage/cohabitation, age, marital status of partner Relationship history Marriage/cohabiting. Works backwards from last

relationship. Seeks details of: dates, sex, age, marital status, marriage, separation/divorce

Partnership status (Based on work undertaken by Dr Ann Berrington, University of Southampton) The following is based on notes made by Dr Berrington whilst exploring the marital status and partnership history data. A number of problems are identified and solution suggested. The data have not been modified in line with these suggested solutions. Users may make such changes if they wish. 1. Proxy interview – IF PROXY=1 For these 66 individuals no information about marital status is available. However, information is available in the variables PROXYAC1, PROXYAC2, PROXYAC3 and PROXYAC4 on whether these cohort members had ever been married, had children, a job or education. Just 5 are reported as having been married. 2. Inconsistencies between legal and de-facto marital status The documentation suggests that at the start of the interview current de-facto marital status is asked for all persons in the household. Later on in the questionnaire, cohort members are asked for their legal marital status. Cross-tabulation of these two variables (MS and MARSTAT2) suggests a number of inconsistencies. a) 140+1 =141 cases give a valid legal marital status but have missing information for de-facto

marital status.

Use the information from the household grid to identify the de-facto marital status (i) MARSTAT2=1, MS=missing (n=76) Most of these people do not have a current partner in the household. i.e. NPART=0, NSPOUSE=0. Code these as MS=3. 4 cases have a current partner – code these as MS=2. However, only one of these four goes on toprovide any information about their current partner (ii)MARSTAT2=2 or 3, MS=missing (n=42) Classify all those living with a spouse (NSPOUSE=1) as de-facto currently married (n=12). The remainder can be classified as de-facto separated (n=30). NOTE: This assumes that all there is no spouse who was missed on the household grid – see comment later on. (iii) MARSTAT2=4 or 5 or 6, MS=missing 3 cases report a current partner in the household grid. 2 of the 3 provide details of this current partner. These 3 can be coded as cohabiting, the remainder can be coded as MS=MARSTSAT2.

b) 12 people gave a valid code for de-facto marital status but no information for legal marital

status.

Code all of those who have a de-facto status of “single never married” as legal marital status “never married” (n=3).


Code those whose de-facto marital status is “divorced” as legal marital status “divorced” (n=1). Code those whose de-facto marital status is “separated” as legal marital status “separated” (n=1). Use information from the partnership history data to identify the legal marital status of the 7 cases who are currently married or cohabiting. I.e. identify whether these individuals have been previously married.

c) 24 cases report themselves as legally separated but de-facto married. 12 appear to have a

partner in the household. We can’t tell for definite whether these are married to their partner as CURPARTB is missing for all: code either as MS=1 or 2. The remaining 12 do not appear to have a partner: code as MS=4.

d) 2 cases say that they are legally separated and say that they are single never married. In fact

they were both previously legally married and experienced marital dissolution. One went on to have a current cohabiting partnership and should be coded as MS=2, whilst the other should be coded MS=4.

e) 10 cases say that they are legally divorced but currently married. 7 of these appear to be

currently cohabiting and could be coded MS=2. The remaining 3 do not have a current partner and so should be coded MS=5.

f) 2 cases say that they are legally widowed and currently married. Both appear to be cohabiting

and so should be coded MS=2. g) 3 cases say that they are legally divorced but currently widowed. One of these is cohabiting

and should be coded MS=2. The other has no partner and should be coded as MS=5. h) 1 case says that they are legally widowed and their de-facto status is divorced. This person

appears to be cohabiting and could be coded MS=2.

3. Inconsistencies between reported current de-facto marital status (MS) and the derived variable indicating whether a spouse or partner is present.

There are 316 cases who are either currently married or cohabiting (MS=1 or 2) where NPART=0 and NSPOUSE=0. It may be that in all of these cases there is a partner living elsewhere. For the analyses discussed below these 316 cases are deleted. A total of 500 of those who report de-facto status separated, divorced or widowed (MS=4-6), also report a current partner in the household. All 500 provide a valid code for the year and month of their current partnership. These cases can be recoded as cohabiting MS=2. 4. Inconsistencies between CURPART and CURPARTC for those who did not live

together before marriage. There appear to be two start dates for the start of the current partnership among those who said that they did not live together before marriage. Was this on purpose – to act as a check? For 203 cases the year as reported in CURPART and CURPARTC was not the same. For some of these cases CURPARTC>CURPART suggesting perhaps a period of co-residence which was not considered “living together”?? In other cases CURPAETC<CURPART which though possible seems unlikely. 5. Years and months of start and end dates of partnerships. The data appear to be very complete with very few missing cases for months of marriage and cohabitation. Given the obvious heaping of the dates, especially of cohabitation, on June (and September to a lesser extent) it would be helpful to clarify whether these months were estimated in the field by the respondent (as per the questionnaire instructions), whether they were estimated in the field by the interviewer, or whether there has been some post-field imputation. As a result of this estimation there are inevitably some inconsistencies between the start of date of cohabitation and marriage. For example among those respondents who are currently married and who lived with their spouse prior to marriage:


88 cases give the same month and year for the start of cohabitation and marriage 30 individuals have a negative period of premarital cohabitation.



Partnership Histories (BCS70 cohort members only) (Based on work undertaken by Dr Ann Berrington, University of Southampton) The following problem cases have been identified by Dr Berrington for BCS70 cases: NB: The identifiers listed below are for the variable BSERIAL – the unique identifier used for

BCS70 cohort members 1. Union histories incomplete - don’t tell us whether they had a previous partnership 04230096 06467062 06499041 10652074 10921001 11033011 15357009 2. Missing all partnership information 01962229 3. Separated/Divorced/Widowed no current partner and no information on previous partnership 00725069 01245013 01695003 03228043 05794018 06129076 07393091 08042081 09855079 11343019 12294077 4. Separated/Divorced/Widowed, MS=2,4,5 No information on current partnership. 00437064 01239038 01563047 01881229 02013058 02178069 02215063 02218089 03310042 03404044 03685084 03940034 04012039 04318022 04356077 04521026 04540003 05152000 05373082 06338031 07019000 07429086 08007052 08734096 09203084 09353042 10126060 10388022 11144040 11847006 13600077 14438031 14507004 14858093 15467063 16349062 17093038 17173063 17537068 17800044 5. Other Inconsistencies in partnership history which could not be resolved 00509063 01949055 02688031 03203014 04203093 04361077 05458083 06631036 07000047 08656021 09784055 10728074 10745000 11528098 13938039 14692092 16486018 18006056



CHILDREN Children Pregnancy history Works backwards from last pregnancy conceived/

fathered. Seeks details of: outcome, name, sex, birth weight, DOB, problems, smoking in pregnancy, other parent, whereabouts, support and circumstances of absent children

Lone parenthood Periods of ≥1 month. Works forward for up to 4 periods. Seeks details of dates and numbers of children

Infertility Ability to have and plans for (more) children Adopted children For up to 4 children. Seeks details of age of child on

adoption and nature of adoption Partner’s children from a previous relationship

Number, whether seen or visit

Children over 16 (NCDS only) Name, age, economic, marital and parent status Absent formerly resident children (NCDS only)

Name, age, whereabaouts, economic, marital and parent status; CM contact and payments for children absent from the houshold, but who were living with CM in 1991

Family activities Things done as a family Demands of parenting Physical/emotional/time demands, worries, closeness of

family Pregnancy Histories (Based on work undertaken by Dr Ann Berrington, University of Southampton and subsequently by the CLS NCDS/BCS70 Team) In analysing the fertility histories, Dr Berrington noted that the data seems generally to be in good shape, if rather cumbersome to work with. However, she did note “…a significant minority who have reported consecutive pregnancies in one confinement…”. As noted above, during the interview, information was gathered for each child conceived in each pregnancy, starting with the most recent and working back through each earlier pregnancy. This means that the birth dates recorded in the most recent pregnancy must be later than those recorded in the second most recent, etc. It also means that within any one pregnancy, the dates recorded for each child conceived should be similar if not the same. Further investigation by the CLS NCDS/BCS70 Team suggests that Dr Berrington was indeed correct to be concerned. For some cohort members, the dates of successive pregnancies appear to be recorded in the variables reserved for multiple outcomes of a single pregnancy. This appears to be true for some 261 cases (154 BCS70 and 107 NCDS), although in some instances the apparent error may reflect the timing of a mix of outcomes, including live births, stillbirths and miscarriages. For some, there may also be a problem with the order of pregnancies which do not appear to start with the most recent and work backward. Following further investigation, a list of detailed updates will be made available, and the dataset will be updated. In the meantime, a list of the cases which currently appear to be effected by this problem is given below.

NB: The identifiers listed below are for the variables NSERIAL and BSERIAL


NCDS 044014L 045002J 045011K 050059U 052008M 053003F 056052M 085010J 087029S 091017J 092206S 093039E 093052W 094028E 099010M 099050Z 100009C 110213J 110244V 120017Q 120080X 120141R 180007A 181038T 183016T 188030P 200075Z 218005B 235011T 280005D 280064W 282113T 285061S 286044X 380041R 380090E 385005P 400075Q 440006V 481001R 500125L 500376N 504034D 506014J 510097Q 512002L 512012P 513033C 513073R 513143L 514012Z 515008Q 517045H 520056F 528014H 528021D 550188U 560008V 610087U 630006D 650034X 650212V 710034D 720027P 750043F 823501Q 825073F 850027R 850032J 860001C 880017H 882021J 882051T 933019C 937020J 950123U 950168T 950290Q 980004E 982053E 984024J 985067J 986100J 986205Y 986250D 986431J X10020X X25101F X32005A X41020X X60006M X66030R X71009E X77014E X77016K X77021B X78086N X82347P Y00087K Y00293N Y01189Y Y20134D Y20154L Y30150J Y30168D Y30169F Y31043N BCS70 00279065 00387068 00530014 00607090 00700040 01065011 01236012 01251089 0125401401266041 01618019 01763001 02294098 02346044 02441045 02454047 02880084 0291402802934081 02939057 03085020 03264047 03366075 03371075 03575030 04046069 0423907304394031 05014068 05036071 05289032 05399086 05408250 05521004 05533031 0554203205585063 05794018 06031073 06072053 06092005 06123024 06145027 06183082 0626603206305077 06466087 06481063 06501030 06570089 06641012 06666041 06707036 0679702206853094 06959022 07052079 07244007 07342034 07416084 07469091 07494043 0805205708078061 08239086 08263063 08283015 08302007 08409011 08445015 08460092 0866707308716094 08787002 08818021 08930049 08962028 08995083 09054086 09145062 0920405909261091 09262066 09310011 09318013 09324089 09490099 09495075 09573049 0959105109716072 09803047 09840026 09891083 10047010 10215086 10455045 10477048 1052704410582025 10657050 10675052 10732099 10937029 11032036 11047089 11162042 1121203811487002 11636000 11706049 12091097 12155070 12434074 12609076 12958013 1318907813561032 13930037 13945090 14137024 14293058 14538008 14671064 14687092 1485809314869044 14978022 15084081 15093082 15440084 15455036 15511007 15527035 1556704015571065 15737066 15810064 16102076 16152057 16224056 16354062 16568094 1663004016717092 16822069 17386019 18013006 18046061 20043000 20091000 20212000 2030500021667000



FAMILY SOCIAL RELATIONSHIPS & SUPPORT Family, Social Relationships & Support Contact with family Parents alive/contact/close/divorce/worries/help. Contact

with siblings/in-laws/grandparents Emotional support Sources/nature of emotional support Review of these data by the CLS NCDS/BCS70 Team and by advisors has, to date, revealed no particular problems, nor have any derived variables been developed. The data will continue to be investigated and, if necessary, detailed updates will be made available, and the dataset will be updated.



FAMILY INCOME Family Income Other Income Income from benefits and regular income form other

sources received by respondent and/or partner Financial situation Organisation of household money and money problems (Based on work undertaken by Dr Gaëlle Pierre, Institute for Employment Research, University of Warwick) In reviewing the data on ‘Family income’, Dr Pierre has reported that, for some BCS70 cases, values for amount of benefit received (BENAMT- BENAMT25) and the amount of other income (INCAMT- INCAMT12) appear quite high. This will be subject to further investigation by the CLS NCDS/BCS70 Team, following which a list of necessary updates will be made available, and the dataset will be updated. In the meantime, users should use these data with some caution.



EMPLOYMENT Employment Economic activity Fairly standard question Current job Seeks details of: earnings, hours, fringe and

Other benefits, pensions, prospects Other paid work Weekly earnings from odd/casual jobs Currently unemployed How became unemployed and job search Labour market histories Periods in a job or not in a job lasting ≥1 month. Works

backwards. Seeks details of: circumstances and dates. If in job gathers details of job, employer, responsibilities (for SOC and SIC coding)

Partner’s job Age left full-time education, economic status, job, employer, responsibilities (for SOC and SIC coding), earnings

Earnings and hours worked (Based on work undertaken by Dr Gaëlle Pierre, Institute for Employment Research, University of Warwick) In reviewing the data on employment, Dr Pierre has noted the following: 1. Some respondents report numbers of hours worked below 30 while reporting f/t employment (for both employees and self-employed). 2. Some of the observations on pay appear very low/high. Further information about earnings are given below. 3. Some of the Standard Industrial Classification (SIC) codes are not in the SIC classification, and some are missing. Further details of the SIC coding are to be found in the following which also accompanies the data deposit:

National Centre for Social Research and Centre for Longitudinal Studies NCDS/BCS70 Teams (2001) National Child Development Study, British Cohort Study 1970: Technical Report

4. Some of the previous activities have the same start dates. 5. Some of the start years for previous activities are equal to 9999 or 9998. As noted above, the following missing values will be found within the data:



6. Some of the start years for the current activity go back to when the respondent was younger than 16. One observation reports a start year of 1958 for a BCS70 respondent.


These problems will be subject to further investigation by the CLS NCDS/BCS70 Team, following which a list of necessary updates will be made available, and the dataset will be updated. In the meantime, users should use these data with some caution.



Earnings (Based on work undertaken by Lorraine Dearden and Alissa Goodman, Institute for Fiscal Studies) In reviewing the data on earnings, Lorraine Dearden and Alissa Goodman have identified a number of problems and produced a number of very valuable revised and derived variables. These are briefly described below. Further details of all variables are to be found in the detailed documentation supplied by Lorraine Dearden and Alissa Goodman, which is included as Appendix 4. Section 1. Interview date 1. Corrected interview date Name: Iintdate Description: Corrected interview date Purpose: To correct interview dates that were originally coded as outside the interviewing

period. In the original data there were 23 individuals whose interview dates were not between 10/99 and 9/00.

Section 2. Deflators 2. RPI at interview date Name: rpi_int Description: Retail price index (RPI) for month of interview, where Jan 2001 = 1.000. Purpose: To allow wages, incomes etc. to be converted into other prices. To uprate wages to

Jan 2001 prices, divide by this deflator. Section 3. Employment variables 3. Total usual weekly hours Name: Hours Description: Total usual weekly hours, including paid and unpaid overtime, not including meal

breaks. Purpose: To derive a measure of total hours usually worked, in order to convert gross and net

pay information available into an hourly pay measure. 4. Imputed/Corrected net last pay Name: Inetpay Description: Imputed or corrected net last pay variable. Purpose: Clean last net pay variable, corrected for implausible values. 5. Imputed/Corrected net period Name: Inetprd Description: Imputed or corrected net last pay PERIOD variable. Purpose: Clean last net pay PERIOD variable, corrected for implausible values and imputed

from gross where net unavailable 6. Imputed/Corrected gross pay Name: Igropay Description: Imputed or corrected gross last pay variable. Purpose: Clean last gross pay variable, corrected for implausible values and imputed from

gross where net unavailable 7. Imputed/Corrected gross period Name: Igroprd Description: Imputed or corrected gross last pay PERIOD variable.


Purpose: Clean last gross pay PERIOD variable, corrected for implausible values and imputed from gross where net unavailable

8. Hourly net last pay Name: hr_net Description: Hourly net last pay. Purpose: Clean hourly equivalent last net pay. Where net pay variables are missing but gross

pay variables are available, net pay has been imputed from gross using known parameters of the tax system in the relevant year.

9. Annual net last pay Name: ann_net Description: Clean annual equivalent net last pay. Where net pay variables are missing but gross

pay variables are available, net pay has been imputed from gross using known parameters of the tax system in the relevant year.

Purpose: Clean annual equivalent last net pay, with missing values imputed wherever possible. 10. Hourly gross last pay Name: hr_gro Description: Clean hourly equivalent last gross pay. Where gross pay variables are missing but

net pay variables are available, gross pay has been imputed from net using known parameters of the tax system in the relevant year.

Purpose: Clean hourly equivalent gross last pay, with missing values imputed wherever possible.

11. Annual gross last pay Name: ann_gro Description: Clean annual equivalent last gross pay. Where gross pay variables are missing but

net pay variables are available, gross pay has been imputed from net using known parameters of the tax system in the relevant year.

Purpose: Clean annual equivalent gross last pay with missing values imputed wherever possible.

Section 4. Partner’s earnings variables 12. Partner’s Imputed/Corrected last net pay Name: Ipnetpay Description: Partner’s last net pay variable, implausible values corrected. Purpose: Clean partner’s last net pay. 13. Partner’s Imputed/Corrected last net pay PERIOD Name: Ipnetprd Description: Partner’s last net pay PERIOD variable, implausible values corrected. Purpose: Clean partner’s last net pay PERIOD. 14. Partner’s Imputed/Corrected last net pay PERIOD Name: Ipnpred Description: Partner’s last net pay PERIOD variable, implausible values corrected. Purpose: Clean partner’s last net pay PERIOD. 15. Partner’s annual last net pay Name: ann_pnet Description: Annual equivalent of partner’s last net pay, implausible values corrected. Purpose: Clean annual equivalent partner’s last net pay 16. Partner’s IMPUTED annual last gross pay Name: ann_pgro Description: Annual equivalent of partner’s last gross pay, imputed from partner’s last net pay

using the known parameters of the tax system in the relevant year.


Purpose: Clean annual equivalent partner’s last gross pay - imputed from net. Section 5. Intermediate variables – used during cleaning 17. Type of fix to pay variable and period codes Name: fixtype Description: Categorical variable containing the type of coding error (if any) which we have

identified in the pay information of each employee, and the fix which has been implemented.

Purpose: To identify and record corrections to last net and gross pay variables. 18. Indicator of odd proportion of gross to net pay Name: oddprop Description: Indicator of whether ratio of gross to net pay suggests coding error in some of the pay

variables which REMAINS UNCORRECTED. Purpose: To identify odd wage information 19. Indicator of wage variable probably too high or too low Name: oddwage Description: Indicator of whether pay variables remain apparently implausibly high or low AFTER

CORRECTIONS have been implemented. Purpose: To identify odd wage information 20. Missing hours variable Name: zhours. Description: Indicator of Usual weekly hours variable (hours) missing. 21. Missing hourly net pay indicator Name: zhr_net. Description: Indicator of if/ why net hourly pay variable is missing. 22. Missing hourly gross pay indicator Name: zhr_gro. Description: Indicator of if/ why gross hourly pay variable is missing. 23. Odd partner’s wage indicator Name: podd. Description: Indicator of apparently implausibly high or low partner’s net wage, remaining AFTER

CORRECTIONS have been implemented.



LIFELONG LEARNING Lifelong Learning Qualifications Age left full-time education and educational and vocational

qualifications held (subjects, grades, dates awarded, where studied)

Current course for qualification Qualification, subject(s), date started, where studied Assessment of current/most recent course

Why taken and expected/experienced benefits

Other courses and training Number, why taken and expected/experienced benefits No formal learning Reasons why no learning Learning overview Useful and enjoyable periods of learning Contact with information technology Use of computers at home and at work Literacy and numeracy Problems with reading, writing and maths, implications and

courses to improve Age left education (Based on work undertaken by: • Dr Gaëlle Pierre, Institute for Employment Research, University of Warwick) • Samantha Parsons, Centre for Longitudinal Studies, Institute of Education, University of

London) In reviewing the data Lifelong Learning, Dr Pierre has noted problems with the data on the age of leaving full-time eduction: “...several answers are >30 for NCDS and some NCDS respondents say "still in f/t education…". Samantha Parsons has reviewed these and related data and produced the revised and derived variables are identified below. Further details of all variables are to be found in the detailed documentation supplied by Samantha Parsons, which is included as Appendix 5. Details of further problems identified by Samantha in relation to other aspects of Lifelong Learning and her suggested solutions are summarised below. Details of problems in other areas of the data are given in the sections dealing with ‘Health’ (Diet, food and exercise; Eating problems; Smoking; Drinking; and Accidents), and with the ‘Self-completion’ (Skills and Illegal drug use). Age left full-time education

variable system-missing user-missing: 8 (^ indicates 98)

user-missing: 9 (^ indicates 99)

valid n

N % N % N % actagel2 607 2.7 ^1 0 22072 actagel2 details age CM left full-time education agelfte2 66 0.3 6 0 36 0.2 22572 agelfte2 captures CMs who were in full-time education at time of interview furthed2 607 2.7 22073 furthed2 details CMs who returned to full-time education less than 3 years after initially leaving full-time education lftmore2 20522 90.5 ^2 2156 lftmore2 details age CM left second period of full-time education Plefted 5981 26.4 ^12 0.1 ^347 2.1 16340 plefted details age CM partner left full-time education


Derived AGE LEFT FULL-TIME EDUCATION Variables



valid n

N % N % N % Ageftedu 109 0.5 22571 6 category variables combining information from actagel2 and agelfte2. Values: 1'before 16' 2'at 16' 3'post 16' 4'post 18' 5'post 21' 6'still in ft edu'. Agelfted 109 0.5 22571 continuous variable combining information from actagel2, agelfte2, furthed2 and lftmore2. Value range: 14 to 42, 99=still in education Ageledu 109 0.5 22571 6 category variable collapsing information in agelfted. Values: 1'before 16' 2'at 16' 3'post 16' 4'post 18' 5'post 21' 6'still in ft edu' plefted1 6340 28.0 16340 5 category variable collapsing information in plefted. Values: 1'before 16' 2'at 16' 3'post 16' 4'post 18' 5'post 21'



Qualifications Problems with reporting of GCSEs (Based on work undertaken by Andrew Jenkins, Institute of Education, University of London) In reviewing the data on qualifications, Andrew Jenkins identified problems with the reporting of GCSE qualifications. The nature of the problem is described below in a summary of the report prepared by Andrew. The full report is included as Appendix 6. 1. The Extent of the Problem The earliest date at which it is possible to have obtain a GCSE qualification is Summer 1988, yet some 94 NCDS cohort members and over 1,000 BCS70 cohort members (CMs) report having obtained one or more GCSEs before 1988. To try to resolve this apparente inconsistency, the data for all of the 94 NCDS CMs and the first 200 BCS CMs have been checked against earlier sweeps of the NCDS/BCS surveys where infoamtion on qualifications is also available. 2. GCSE Date Errors for the NCDS In the 2000 sweep of the data, those NCDS CMs who had participated in the previous sweep in 1991 were asked to supply information on all qualifications which they had obtained since 1991; those who were absent from the previous sweep were asked about all qualifications which they had obtained since 1974, or school leaving age. Asking for information over such a long spell of time is very likely to introduce inaccuracies due to imperfect recall. However, for NCDS, the 2000 sweep represents the sixth follow-up on the lives of cohort members, which means that there is potential information in earlier sweeps which can be used use for checking. In 1978, schools and colleges attended by cohort members were contacted and asked to supply details of the qualifications which NCDS CMs had obtained. This is a valuable resource, since it represents an objective source of information on qualifications obtained. The main drawback is that it only covers the period up to 1978. In what follows it is referred to as the EXAMs file. It contains data on the number of qualifications obtained, and also some data on the subjects of qualifications. NCDS4 (1981) some details of qualifications obtained since 1974, but NCDS5 (1991) is more informative. In NCDS5 CMs were asked (a) about qualifications ever obtained and (b) about qualifications obtained since 1981. Note that the NCDS questionnaire did ask for GCSEs, as well as O levels and CSEs. The main drawback of NCDS5 is that it does not tell us about the number of qualifications. For example, CMs were asked if they had CSEs at grades 2-5, but were not asked to report how many they had obtained. It also, of course, has problems of recall errors. The EXAMS file, supplemented by NCDS5, then, are the main sources which we can use to check the accuracy of the data in NCDS6 (2000). A detailed case-by-case listing for each of the 94 NCDS records in question is contained in Appendix ?, and this should be referred to when considering the summary given here. Fortunately, most of the 94 cases were present in the EXAMS file, and since the majority of erroneous cases were reporting that they got their GCSEs in the 1973 to 1976 period, the data contained in the EXAMS file is relevant. In more detail, all but ten of the 94 cases were present in the EXAMs file. Out of these, five were present in NCDS5. There is no earlier data for the remaining five. For some of the remaining cases, EXAMS file data was found to be not relevant because they were reporting GCSE qualifications after 1978, but fortunately this applied to 7 cases. Three of these were covered by NCDS5, but the remaining 4 were not. In sum, then, there were only 9 cases where data could not be obtained from earlier sweeps to check the GCSE errors in NCDS6. The majority of those for whom data is available would appear to have CSEs rather than the GCSEs reported in NCDS6. However, some have O levels, and some a mixture of O levels and CSEs. Although a good many cases appear to show confusion between GCSEs and either O levels or CSEs, some cases are less simple, showing differences in the number of O levels recorded in the


EXAMS file and the number of GCSEs listed in NCDS6. In other cases, there is little evidence in the EXAMS file that the qualifications reported in NCDS6 were actually obtained. Other cases show duplication, and can be easily resolved out in the absence of EXAM data. 3. GCSE date errors among a sample of BCS70 cases Because there are such a large number of GCSE errors among the BCS70 cohort, it has only been possible to look at the first 200 of these (rather less than a fifth of the total). The BCS70 CMs were one of the last cohorts to sit O levels and CSEs as their school-leaving exams before they were replaced by GCSEs. BCS CMs should only have GCSEs if they stayed in education to at least age 18, or subsequently returned to study for a qualification. The only source of data for checking the responses in BCS 2000 is the earlier BCS postal survey conducted in 1996. Questions about qualifications in the BCS 1996 survey were asked in the order: CSE, O level, GCSE, rather than the order GCSE, O level, CSE in BCS 2000. Thus checking BCS70 2000 against BCS70 1996 should help to root out errors which may have arisen from the order in which the questions were asked in BCS70 2000. However, BCS70 1996 had a relatively small sample size (n = 9,003), so many of those in BCS70 2000 may not be present in BCS 1996. Of the 200 BCS70 CMs checked, some 134 (67%) were also present in BCS70 1996. Out of these, 2 also reported GCSEs in BCS70 1996 (although not necessarily the same number as in BCS70 2000); 17 provided no information about qualifications in BCS70 96, and 75 did not report any GCSEs in BCS70 96. Detailed information on each case is presented in Appendix ?. Among the 134 for whom 1996 data are available it is usually possible to make some reasonable guess as to what should be done about the erroneous GCSE data in BCS 2000. For many, it is clear that the GCSE data should be recoded to O levels, while in others the GCSEs should be recoded as CSEs. Other cases are not quite so obvious, and some, where there are large discrepancies between 1996 and 2000, or where errors have been repeated. The real difficulty is, however, that we have a substantial number of cases not present in 1996, so there is no data to check the BCS 2000 results against. 4. O levels and CSEs reported as gained after 1988 Among all the NCDS CMs, some 59 had reported obtaining one or more O levels after 1988 - after they were (presumably) phased out in 1988. The dates for the reported O levels were all in the 1990s. The source of the error is likely to be either that (a) they have actually got GCSEs rather than O levels or (b) a mistake was made inputting the data, so that the O level was actually obtained in,say, 1974 but this was coded as 1994. Some 47 of the 59 cases (79%) were also present in NCDS5, which means that they were (or should have been) only asked questions about qualifications obtained since 1991. This would make it unlikely that error (b) occurred. The data do indeed show that most of these 47 were only reporting O levels obtained in the 1990s: 28 reported a single O level obtained in the 1990s and no earlier O levels; a further 13 reported two O levels obtained in the 1990s and no O levels. However, the remaining six of the 47, despite being present in NCDS5 in 1991, also reported O levels obtained in both the 1970s and the 1990s. For the 12 not present in 1991, it is difficult to discriminate between the (a) and (b) errors listed above. So it is not possible to resolve all the cases, but for many recoding to GCSE would be a sensible solution. Only two NCDS CMs reported obtaining CSEs after 1988.


For BCS, there appear to be 133 cases in which CMs report O levels obtained after 1988, and 32 cases report CSEs obtained after 1988, but these have not been studied in any detail.



Highest qualification (Based on work undertaken by:

• Andrew Jenkins, Institute of Education, University of London and Professor Gerald Makepeace, University of Cardiff

• Samantha Parsons, Centre for Longitudinal Studies, Institute of Education, University

of London) (Based on work undertaken by Andrew Jenkins, Institute of Education, University of London and Professor Gerald Makepeace, University of Cardiff In reviewing the data relating to qualifications, Andrew Jenkins and Gerald makepeace have developed a number of derived variables relating to qualifications. In particular, they provide dereived variables for the highest academic qualification, the highest vocational qualification and the highest NVQ level recorded in the 2000 survey, as shown below. HIACA00 'Highest academic qualification recorded in 2000 survey'.

-9 Missing 0 'None ' 1 ' Bad O-levels' 2 'CSE 2-5, other Scottish school qualification' 3 'Good O levels' 4 '1 A level or more than 1 AS level at grade A-C' 5 '2 or more A-levels' 6 'Diploma' 7 'Degree, PGCE, other degree level' 8 'Higher degree ' .

NVQACA00 'Highest NVQ level from an academic qualification in 2000 survey'.

GHMACA00 'Highest academic qualification in 2000 survey ghm measure see also IFS'. 0 'None' 1 'Bad O levels, CSE 2-5' 2 'Good O-levels, 1 A-level' 3 '2 or more A-levels' 4 'Sub-degree' 5 'Degree' 6 'Higher Degree' .

HIVOC00 'Highest vocational qualification recorded in 2000 survey'.

-9 Missing 0 'None ' 1 'C&G Opo & Other, RSA 1, HGV, Other tech or bus qualification' 2 'C&G Craft-Inter-Ord-Part1-Cant say, RSA2, Insig,JIB NJC etc' 3 'C&G Advanced-Final-Part II or III, ONC-OND, RSA 3 ' 4 'Nursing, BTEC HNC-HND, HNC-HND, C&G Full Tech (FTC)' 5 'NVQ awarded at level 5'.

HiNVQ00 'Highest NVQ level (whether academic or vocational)'


It is important to stress that:

• These variables are based only on the data gathered in the most recent survey.

• For the majority of NCDS cohort members, the variables may have only limited value as they only relate to qualifications gained since the age of 33 years. This is because for most NCDS cohort members, information was only gathered on qualifications gained since the last follow-up in 1991. Only if they did not take part in the 1991 survey was information gathered about qualifications gained since 1981 – the date of the previous follow-up. In this case, the variables relate to qualifications gained since the age of 23. Comprehensive information about qualifications gained since the age of 16 can only be obtainedby linking the data gathered in the new survey with that gathered in earler follow-ups.

• For BCS70 cohort members, the variables relate to qualifications gained since the age of 16.

• Previous published work has used slightly different definitions of these variables so some

users may need to adapt the code they provide. They table below shows how the educational qualifications obtained were mapped into the national qualifications framework.

Level of Qualificat

ion

General (Academic)

Vocationally-related (Applied)

Occupational (Vocational)

5 Higher Degree NVQ level 5 PGCE

4 Degree

HE Diploma

BTEC Higher Certificate/Diploma

HNC/HND

NVQ level 4 Professional degree level

qualifications Nursing/paramedic

Other teacher training qualification

City & Guilds Part 4/Career Ext/Full Tech

RSA Higher Diploma 3 A level

AS levels Scottish Highers

Scottish Cert of 6th Year Studies

Advanced GNVQ BTEC National Diploma

ONC/OND

NVQ level 3 City & Guilds Part

3/Final/Advanced Craft RSA Advanced Diploma

Pitmans level 3 2 GCSE grade A*-C

O levels grade A-C O levels grade D-E

CSE grade 1 Scottish standard

grades 1-3 Scottish lower or ordinary grades

Intermediate GNVQ BTEC First Certificate BTEC First Diploma

NVQ level 2 Apprenticeships

City & Guilds Part 2/Craft/Intermediate

City & Guilds Part 1/Other RSA First Diploma

Pitmans level 2

1 GCSE grade D-G CSEs grades 2-5 Scottish standard

grades 4-5 Other Scottish school

qualification

Foundation GNVQ Other GNVQ

NVQ level 1 Other NVQ

Units towards NVQ RSA Cert/Other Pitmans level 1

Other vocational qualifications HGV


Jenkins and Makepeace also define a set of binary variables showing whether an individual has a particular qualification. These variables may be used to devise various typologies for training. They are listed below. A copy of the SPSS code and documentation supplied by Jenkins and Makepeace is included as Appendix 7.

Queries about these variables should be directed to the User Support Group ([email protected])


Qualification variables derived by Jenkins and Makepeace

Variable Label GCSEHI00 Has CM obtained one or more GCSEs at grades A to C ? GCSELO00 Has CM obtained one or more GCSEs at grades D to E ? OLEVHI00 Has CM obtained one or more O levels at grades A to C? OLEVLO00 Has CM obtained one or more O levels at grades D to E? CSEHI00 Has CM obtained one or more CSEs at grade 1? CSELO00 Has CM obtained one or more CSEs at grades 2 to 5? ASLHI00 Has CM obtained one or more AS levels at grades A to C? ASLLO00 Has CM obtained one or more AS levels at grades D to G? NASLEV00 Number of AS level passes at grade A-C ALEVHI00 Has CM obtained one or more A levels at grades A to C? ALEVLO00 Has CM obtained one or more A levels at grades D to E? ALEV00 Has CM obtained any A level passes at any grade? NALEV00 Number of A level passes at any grade SCOTLO00 Has CM obtained Scottish SCE standard grades 4-5 or equivalent? SCOT200 Has CM obtained Scottish SCE standard grades 1-3 or equivalent? SCOT300 Has CM obtained Scottish SUPE-SLC lower or ordinary grade? SCOT400 Has CM obtained Scottish SCE-SUPE-SLC higher grade or equivalent? SCOT500 Has CM obtained Scottish Certificate of 6th Year Studies? SCOT600 Has CM obtained other Scottish school qualification? SCOTLO00 Has CM obtained Scottish SCE standard grades 4-5 or equivalent? SCOT200 Has CM obtained Scottish SCE standard grades 1-3 or equivalent? SCOT300 Has CM obtained Scottish SUPE-SLC lower or ordinary grade? SCOT400 Has CM obtained Scottish SCE-SUPE-SLC higher grade or equivalent? SCOT500 Has CM obtained Scottish Certificate of 6th Year Studies? SCOT600 Has CM obtained other Scottish school qualification? HDEG00 Has CM obtained a higher degree? DIP00 Diploma-Certificate-Teacher Training Qual (not pgce) in Higher Ed NURSE00 Has CM obtained any nursing or paramedic qualifications? PGCE00 Has CM obtained a PGCE or other postgraduate certificate in HE? ONCOND00 Does CM have ONC or OND?. HNCHND00 Does CM have HNC or HND?. BTEC100 Does CM have a BTEC etc First-General Certificate? BTEC200 Does CM have a BTEC etc First-General Diploma? BTEC300 Does CM have a BTEC etc National Certificate Diploma? BTEC400 Does CM have a BTEC etc Higher Certificate Diploma? BTEC500 Does CM have other BTEC qualification? CITY100 Does CM have a City and Guilds Part 1 CITY200 Does CM have a City and Guilds Part2 or Craft or Intermediate CITY300 Does CM have a City and Guilds Part3 or Final or Advanced Craft CITY400 Does CM have a City and Guilds Part4 or Career Extension or Full tec CITY500 Does CM have a City and Guilds Other C&G qualification RSA100 Does CM have RSA certificate? RSA200 Does CM have RSA First Diploma? RSA300 Does CM have RSA Advanced Diploma or Certificate? RSA400 Does CM have RSA Higher Diploma RSA500 Does CM have other RSA qualification? PIT100 Does CM have a Pitmans level 1? PIT200 Does CM have a Pitmans level 2?


Variable Label PIT300 Does CM have a Pitmans level 3? PIT400 Does CM have otherPitmans qualification? NVQ100 Does CM have an NVQ at Level 1? NVQ200 Does CM have an NVQ at Level 2? NVQ300 Does CM have an NVQ at Level 3? NVQ400 Does CM have an NVQ at Level 4? NVQ500 Does CM have an NVQ at Level 5? NVQ600 Does CM have an NVQ at Level 6? NVQ700 Does CM have Trusts towards NVQ-SVQ? NVQ800 Does CM have other NVQ?. GNVQ100 Does CM have GNVQ Foundation? GNVQ200 Does CM have GNVQ Intermediate? GNVQ300 Does CM have GNVQ Advanced? GNVQ400 Does CM have other GNVQ qualification? APPREN00 Has CM obtained any recognised trade apprentice qualifications? HGV00 Has CM obtained at least one hgv qualification? OTHVOC00 Has CM obtained at least one other vocational qualification? ACAMISS Dummy for missing value for do you have an educational qualification VOCMISS Dummy for missing value for do you have a vocational qualification ALLMISS Missing values, Do you have any of these qualifications? PROXY Was interview completed by a proxy YTS00 Has CM done a YTS course?. GOV00 Has CM done another gov course (including New Deal)? MODAPP00 Has CM done a modern apprenticeship? NFAIL00 Number of courses started where CM did not obtain the qualification ACCESS00 Has CM done an access course? WORKC00 Number of other work related courses LEIS00 Number of courses for interest or leisure READ00 Number of courses to improve reading WRITE00 Number of courses to improve writing MATHS00 Number of courses to improve maths ACA000 1 if CM has no academic qualifications ACA100 1 if CM has bad O-levels . ACA200 1 if CM has CSE 2-5, Other Scots ACA300 1 if CM has good O-levels . ACA400 1 if CM has 1 A-level or more than 1 AS at grade A-C ACA500 1 if CM has 2 or more A-levels ACA600 1 if CM has diploma . ACA700 1 if CM has degree, PGCE, other degree level ACA800 1 if CM has higher degree. VOC000 1 if CM has no vocational qualifications VOC100 1 if CM has vocational qualification equivalent to NVQ1 VOC200 1 if CM has vocational qualification equivalent to NVQ2 VOC300 1 if CM has vocational qualification equivalent to NVQ3 VOC400 1 if CM has vocational qualification equivalent to NVQ4 VOC500 1 if CM has vocational qualification equivalent to NVQ5 HIACA00 Highest academic qualification recorded in 2000 survey NVQACA00 Highest NVQ level from an academic qualification in 2000 survey GHMACA00 Highest academic qualification in 2000 survey ghm measures see also IFS HIVOC00 Highest vocational qualification recorded in 2000 survey


Variable Label HINVQ00 Highest NVQ level (whether academic or vocational) GROUP 3 categories - BCS70 and NCDS by whether present in 1991 survey

(Based on work undertaken by Samantha Parsons, Centre for Longitudinal Studies, Institute of Education, University of London) Building on the work of Jenkins and Makepeace, Samantha Parsons, has developed derived variables which match the measures of highest academic and vocational qualifications, but which incorporate information gathered during the NCDS5 follow-up in 1991. This ensures that they provide information based on all qualifications gained since the age of 16. The variables are listed below. It is important to stress that in these variables:

• Uses the NCDS5 variables: N501441 to N501469

• Degree is now coded at NVQ4 level.

• At NCDS5 it is impossible to tell how many ‘A' levels cohort members have. Hence, this code assumes NCDS5 ‘A’ levels are 2 or more. The Jenkins and Makepeace variables differentiate between those with 1 ‘A’ level, and those with 2 or more.

• City & Guilds Full Technological is allocated to NVQ4. The Jenkins and Makepeace variables

allocate this qualification to NVQ3.

HQUAl33 "Highest qual gained at age 33 – based on NCDS5 variables".

0 ‘No qualification’ 1 ‘CSE 2-5/equiv NVQ1’ 2 ‘O Level/equiv NVQ2’ 3 ‘A Level/equiv NVQ3’ 4 ‘Higher qual NVQ4’ 5 ‘Degree/higher NVQ5/6’ -1 ‘No information’.

HQ332K 'highest NVQ qualification - academic or vocational - recoding 2000 to match NCDS5'. HACAD33B 'ncds33: highest academic qualification to match HIACA00'.

0'no quals' 2'cse2-5' 3'good o levels' 5'a levels' 6'diploma' 7'degree PGCE' 8'higher degree'.

HACNVQ33 'ncds33: highest NVQ level qual from academic qualification to match NVQACA00'.


HACGHM33 'ncds33: highest academic qualification - ghm measure to match GHMACA00'. 0'no quals' 1'cse2-5' 2'o levels' 3'a levels' 4'sub degree' 5'degree' 6'higher degree'.

HVOC33 'ncds33: highest vocational qualification - to match HIVOC00'. HQ33A 'ncds33: highest NVQ qualification - academic or vocational - to match HINVQ00'. HQACA00 'highest academic qualification - combining hacad33b and hiaca00'.

0'no quals' 1'bad o levels' 2'cse2-5' 3'good o levels' 4'as levels or 1 a level' 5'a levels' 6'diploma' 7'degree PGCE' 8'higher degree'.

HQANVQ00 'highest NVQ level qual from academic qualification - combining hacnvq33 and nvqaca00'. HQAGHM00 'highest academic qualification - ghm measure - combining hacghm33 and ghmaca00'.

0'no quals' 1'cse2-5' 2'o levels' 3'a levels' 4'sub degree' 5'degree' 6'higher degree'.

HQVOC00 'highest vocational qualification - combining hvoc33 and hivoc00'. HQNVQ00 'highest NVQ qualification - academic or vocational - combining hq33a and hinvq00'.

A copy of the SPSS code supplied by Samantha Parsons is also included as Appendix 7.



Basic skills (Based on work undertaken by Samantha Parsons, Centre for Longitudinal Studies, Institute of Education, University of London) Whilst reviewing the data on basic skills, Samantha Parsons identified a number of problems and produced a number of very valuable revised and derived variables. These relate to the following:

1. Maths difficulties 2. Reading difficulties 3. Writing difficulties 4. Difficulties resulting from having 3R problem 5. Course attendance 6. Computer use at home and work

The problems and revised and derived variables are identified below. Further details of all variables are to be found in the detailed documentation supplied by Samantha Parsons, which is included as Appendix 5. Details of further problems identified by Samantha in other areas, and her suggested solutions are given in the sections dealing with ‘Health’ (Diet, food and exercise; Eating problems; Smoking; Drinking; and Accidents), and with the ‘Self-completion’ (Skills and Illegal drug use).


1. Maths difficulties



valid n

n % N % n % mathsprb 66 0.3 4 0 21 0.1 22589 549 (2.4%) had difficulty/were not able to work out change from £10. However, only 426 (78% of 549) asked additional 5 questions on type of maths difficulties CM had: mprbtype 426 addup 426 subtract 426 multiply 1 0 425 subtract 1 0 425 mathcors 66 0.3 4 0 20 0.1 22590 429 (1.9%) had done a course to improve their maths. 429 asked if doing course now or previously mathsnow 22251 98.1 429 mathimp 139 0.6 3 0 22538 6234 (27.7%) ever wanted to improve their maths. 6234 asked multi-coded questions i) why they wanted to improve their maths and ii) where/how they would improve them mthlike1 – mthlike5

16446 72.5 2 6232

mthplac1-mthplac9

16446 72.5 ^4 6230

datesprb 1377 6.1 1 0 15 0.1 21271 696 (3.3%) had difficulty/were not able to work out dates/use a calendar mathconf 11421 50.4 3 0 20 0.1 11236 1536 (13.7%) not confident helping their child(ren) with maths mathskid 11421 50.4 3 0 16 0.1 11240 1093 (9.7%) did not help their child(ren) with maths partmath 12733 56.1 4 0 22 0.1 9921 1439 (14.5%) of CM partner did not help their child(ren) with maths


Derived MATHS DIFFICULTIES variables



valid n

n % n % n % bsamd 1394 6.1 21284 4 category variable combining information given in mathsprb and datesprb. Only for CMs who answered both questions. CM reporting ‘difficulties’ or ‘no, can’t do’ treated as one category. Values: 1’no difficulties’ 2’maths difficulties’ 3’date/calendar difficulties’ 4’maths and 3’date/calendar difficulties’. bsamp1 22254 98.1 426 6 category variable counting the specific maths problems of CMs who had reported maths difficulties in mathsprb. Included mprbtype (recognising numbers), addup, subtract, multiply, and divide. Values: 0’not these problems’ 1’one problem’ 2’two problems’ 3’three problems’ 4’four problems’ 5’all five problems’. bsamp2 22254 98.1 426 4 category variable counting the specific maths problems of CMs who had reported maths difficulties in mathsprb .Included addup, subtract, multiply, and divide. Values: 0’not these problems’ 1’one problem’ 2’two problems’ 3’three problems’ 4’all four problems’. Derived variables from multi-coded mthlike1-mthlike5: reasons why CM wants to improve their maths: for their children, to get a job, to get a promotion, get a better job, for self satisfaction. impmkid impmgjob impmprom impmbjob impmself

16448 72.5 6232

Derived variables from multi-coded mthplac1-mthplac9: means by which CM would improve their maths: on a day college course, on an evening college course, on a weekend college course, on ANY college course, at a community centre, by using local library resources, by using PC packages, by using programmes on TV, by using programmes on the radio, by using books at home. domcolld domcolle domcollw domcoll domcomc domllib dompc domtv domradio dombook

16450 72.5 6230


2. Reading difficulties



valid n

n % n % n % readprb1 1377 6.1 1 0 21302 679 (3.2%) had difficulty/were not able read/understand magazine/newspaper text. 680 asked additional 2 questions on other reading difficulties (includes CM with value of 9 in readprb1) readprb2 22000 97.0 680 readprb3 22003 97.0 3 0 677 readprb4 1377 6.1 3 0 21300 Asks if CM feels ability to read paperwork has improved over the last decade. 789 (3.7%) feel it has got worse. Readcors 66 0.3 4 0 20 0.1 22590 197 (0.9%) had done a course to improve their reading. 197 asked if doing course now or previously Readnow 22483 99.1 197 Readimp 128 0.6 2 0 22550 1856 (8.2%) ever wanted to improve their reading. 1856 asked multi-coded questions i) why they wanted to improve their reading and ii) where/how they would improve them redlike1 – redlike5

20824 91.8 1856

redplac1-redplac9 (redplac8 and redplac9 hold no information)

20824 91.8 ^2 1854

Kidrdcnf 11421 50.4 3 0 21 0.1 11235 561 (5.0%) not confident helping their child(ren) with reading Readkid 11421 50.4 3 0 15 0.1 11241 822 (7.3%) did not read to their child(ren) Partread 12733 56.1 4 0 21 0.1 9922 1271 (12.8%) of CM partner did not read to their child(ren)


Derived READING DIFFICULTIES variables



valid n

n % n % n % bsaread1 1378 6.1 21302 4 category variable combining information given in readprb1, readprb2 and readprb3. CM reporting ‘difficulties’ or ‘no, can’t do’ treated as one category. Values: 1’no reading difficulties’ 2’I type of reading difficulty’ 3’2 types of reading difficulties’ 4’3 types of reading difficulties’. bsaread2 22003 97.0 677 3 category variable combining information given in readprb1, readprb2 and readprb3 for CMs who reported ‘difficulties’ or ‘no, can’t do’ in readprb1 and answered readprb1, readprb2 and readprb3. CM reporting ‘difficulties’ or ‘no, can’t do’ treated as one category. Values: 1’no reading difficulties’ 2’I type of reading difficulty’ 3’2 types of reading difficulties’ 4’3 types of reading difficulties’. Derived variables from multi-coded redlike1-redlike5: reasons why CM wants to improve their reading: for their children, to get a job, to get a promotion, get a better job, for self satisfaction. imprkid imprgjob imprprom imprbjob imprself

20824 91.8 1856

Derived variables from multi-coded redplac1-redplac9: means by which CM would improve their reading: on a day college course, on an evening college course, on a weekend college course, on ANY college course, at a community centre, by using local library resources, by using PC packages, by using programmes on TV, by using programmes on the radio, by using books at home. dorcolld dorcolle dorcollw dorcoll dorcomc dorllib dorpc dortv dorradio dorbook

20826 91.8 1854


3. Writing difficulties



valid n

n % n % n % writeprb 1377 6.1 1 0 21302 1087 (3.2%) had difficulty/were not able to write a thank-you letter to a friend. 1087 asked additional question on spelling difficulties (includes CM with value of 9 in writeprb) wprbtype 21593 95.2 1 0.1 1086 70 (6.4%) never tried to write a thank-you letter to a friend. 1017 asked additional questions on handwriting legibility and difficulties putting down in words what want to say hwritprb 21663 95.5 1 0.1 1016 wordsprb 21663 95.5 1017 writcors 66 0.3 4 0 21 0.1 22589 315 (1.4%) had done a course to improve their writing. 315 asked if doing course now or previously writenow 22365 98.6 315 writimp 138 0.6 3 0 22539 2636 (11.7%) ever wanted to improve their writing. 2636 asked multi-coded questions i) why they wanted to improve their reading and ii) where/how they would improve them wrilike1 – wrilike5

20044 88.4 1 0 2635

wriplac1-wriplac9

20044 88.4 ^5 2631

writconf 11421 50.4 4 0 26 0.1 11229 806 (7.2%) not confident helping their child(ren) with writing writekid 11421 50.4 3 0 14 0.1 11242 1440 (12.8%) did not help their child(ren) with writing partwrit 12733 56.1 4 0 21 0.1 9922 1687 (17.0%) of CM partner did help their child(ren) with writing


Derived WRITING DIFFICULTIES variables



valid n

n % n % n % bsawrit1 1379 6.1 21301 8 category variable combining information given in writeprb, hwritprb, wprbtype and wordsprb. CM reporting ‘difficulties’ or ‘no, can’t do’ treated as one category. Values: 0'no probs' 1'writing spelling handwriting articulating' 2'writing spelling handwriting' 3'writing spelling articulating' 4'writing handwriting articulating' 5'writing spelling' 6'writing articulating' 7'writing handwriting' 8'writing only'. bsawrit2 21594 95.2 1086 4 category variable counting the specific writing problems of CMs who had reported ‘difficulties’ or ‘no, can’t do’ in writeprb. Included writeprb, wprbtype (spelling), hwritprb (handwriting), and wordsprb (articulating). CM reporting ‘difficulties’ or ‘no, can’t do ’in writeprb treated as one category. Values: 1’1 difficulty’ 2’2 types of writing difficulties’ 3’3 types of writing difficulties’ 4’4 types of writing difficulties’. Derived variables from multi-coded redlike1-redlike5: reasons why CM wants to improve their reading: for their children, to get a job, to get a promotion, get a better job, for self satisfaction. Values: 0'not this reason' 1'yes, for this reason'. impwkid impwgjob impwprom impwbjob impwself

20045 88.4 2635

Derived variables from multi-coded redplac1-redplac9: means by which CM would improve their reading: on a day college course, on an evening college course, on a weekend college course, on ANY college course, at a community centre, by using local library resources, by using PC packages, by using programmes on TV, by using programmes on the radio, by using books at home. Values: 0'not using this method' 1'this method'. dowcolld dowcolle dowcollw dowcoll dowcomc dowllib dowpc dowtv dowradio dowbook

20049 88.4 2631


4. Difficulties resulting from having basic skills problems

Variable system-missing user-missing: 8 (^ indicates 98)


valid n

n % n % n % Getjob 21047 92.8 21 0.1 1612 620 (38.5%) CMs with 3R problem felt affected them getting new job Copejob 21047 92.8 15 0.1 1618 391 (24.2%) CMs with 3R problem felt affected them coping in current job Gtpromot 21047 92.8 38 0.2 1595 570 (35.7%) CMs with 3R problem felt affected them getting a promotion Copehmbs 21047 92.8 6 0 1627 253 (15.6%) CMs with 3R problem felt affected them in managing household business Helpkids 21047 92.8 3 0 1630 419 (25.7%) CMs with 3R problem felt affected them being able to help their children lear. In addition, 409 (25.1%) never helped their children to read or learn Copeleis 21047 92.8 4 0 1629 341 (20.9%) CMs with 3R problem felt affected them pursuing other interests


5. Course attendance



valid n

N % n % n % Failqual 66 0.3 3 34 0.2 22577 3404 (8.8%) failed/cot completed course. 3404 asked how many courses they had failed/not completed Numfqual 19276 84.9 ^3 3401 yts (bcs70 only) 11454 50.5 3 12 0.1 11211 3303 (29.5%) CMs had done a yts course. 3303 asked how many they had done, and if they were on a course now Numyts 19377 85.4 ^1 ^1 3301 Ytsnow 19377 85.4 1 3302 Othgov 66 0.3 4 22 0.1 22588 550 (2.4%) CMs had done an ‘other’ type of Government course. 550 asked how many they had done, and if they were on a course now Numgov 22130 97.6 ^3 0.5 547 Govnow 22130 97.6 550 Aptrain 66 0.3 4 24 0.1 22586 329 (1.5 %) CMs had done a modern apprenticeship course. 329 asked how many they had done, and if they were on a course now Numap 22351 98.5 329 Apnow 22351 98.5 329 Actrain 66 0.3 4 23 0.1 22587 352 (1.6%) CMs had done an access course. 3303 asked how many they had done, and if they were on a course now Numac 22328 98.4 352 Acnow 22328 98.4 352 Wrktrain 66 0.3 4 24 0.1 22586 7871 (34.7%) CMs had done a work-related training course. 7871 asked how many they had done, and if they were on a course now Numwrktr 14809 65.3 ^66 0.8 7805 Wrktrnow 14809 65.3 1 0 7870 Leiscors 66 0.3 5 23 0.1 22586 5528 (24.4%) CMs had done a course for interest or leisure. 5528 asked how many they had done, and if they were on a course now Numleis 17152 75.6 ^4 0.1 5524 Leisnow 17152 75.6 5528


Derived COURSE ATTENDANCE variables



valid n

N % n % n % yts1 11470 50.6 11210 3 category variable combining information from yts and ytsnow. Values: 0'no' 1'yes, previously' 2'yes, now'. Gov1 94 0.4 22588 3 category variable combining information from othgov and govnow. Values: 0'no' 1'yes, previously' 2'yes, now'. apt1 92 0.4 22586 3 category variable combining information from aptrain and apnow. Values: 0'no' 1'yes, previously' 2'yes, now'. access1 95 0.4 22587 3 category variable combining information from actrain and acnow. Values: 0'no' 1'yes, previously' 2'yes, now'. wrt1 93 0.4 22585 3 category variable combining information from wrktrain and wrktrnow. Values: 0'no' 1'yes, previously' 2'yes, now'. Leis1 94 0.4 22586 3 category variable combining information from leiscors and leisnow. Values: 0'no' 1'yes, previously' 2'yes, now'. read1 90 0.4 22590 3 category variable combining information from readcors and readnow. Values: 0'no' 1'yes, previously' 2'yes, now'. Writ1 91 0.4 22589 3 category variable combining information from writcors and writenow. Values: 0'no' 1'yes, previously' 2'yes, now'. math1 90 0.4 22590 3 category variable combining information from mathcors and mathsnow. Values: 0'no' 1'yes, previously' 2'yes, now'.


6. Computer use at Home and Work



valid n

n % n % n % pchome 66 0.3 4 20 0.1 22590 13379 (59.0%) CM had a PC at home. 13379 asked how often they used the hpcuse 9301 41.0 4 0 13375 11289 (84.4%) CMs used the PC at home. 11289 asked about the ways they use a PC howuse01-howuse33 (howuse18-howuse33 hold no information)

11391 50.2 ^8 0.1 11281

pcwork 3915 17.3 2 9 0 18754 12026 (64.1%) CM use a PC at work. 12026 asked how often use a PC and the ways they use a PC wpcuse 10654 47.0 1 12025 howuse34-howuse46

10654 47.0 ^2 0 12024

Derived HOME COMPUTER variables



valid n

n % n % n % pchome1 94 0.4 22586 6 category variable combining information from pchome and hpcuse. Values: 0'no' 1'yes, but do not use it' 2'yes, use it less than once a week' 3'yes, use it once a week' 4'yes, use it 2-4 times a week' 5'yes, use it daily' Derived variables from multi-coded howuse01-howuse33. For 11289 (50.0%) CMs who have and use a home computer. Ways CM uses home computer: for word processing, WWW, e-mail, data analysis, data base work, design packages, playing games, sending/receiving faxes, using CD ROM or Encyclopedia, composing music, listening to music, photography, programming, managing home finances, spreadsheets, webb design, scanning, other things, unspecified things. Values: 0'not use home PC this way' 1'yes, uses home PC this way'. hwp hwww hemail hdatan hdatab hdesign hgames hfax hencyrom hmusicc hmusicl hphoto hprog hhomefin hspread hwebdes hscan hother hunspec

11391 50.2 11289


Derived WORK COMPUTER variables



valid n

n % N % n % pcwork1 3927 17.3 18753 5 category variable combining information from pcwork and wpcuse. Values 0'no' 1'yes, use it less than once a week' 2'yes, use it once a week' 3'yes, use it 2-4 times a week' 4'yes, use it daily'. Derived variables from multi-coded howuse34-howuse46. For 12025 (64.1%) CMs who use a PC at work. Ways CM uses work computer: for word processing, WWW, e-mail, data analysis, data base work, design packages, playing games, sending/receiving faxes, using CD ROM or Encyclopedia, composing music, listening to music, photography, programming, spreadsheets, webb design, scanning, other things, unspecified things. Values: 0'not use work PC this way' 1'yes, uses work PC this way'. wwp wwww wemail wdatan wdatab wdesign wgames wfax wencyrom wmusicc wmusicl wphoto wprog wother

10655 47.0 12025



HEALTH Health General health Self-assessed health and experience of list of conditions,

including age at onset and contact with doctor Long-term health conditions Details of longstanding illnesses, etc (including limiting

impact and age at onset), impact on employment, registered disabled.

Respiratory problems Coughing, phlegm and shortness of breath Mental health Experience of mental health problems, including age at

onset Seeing and hearing Problems with sight/eyes and with hearing Other conditions Details of other health conditions requiring regular medical

supervision Accidents/injuries Works backwards. Details of accidents/ injuries/assaults

(age, why admitted, out/in-patient, type of injury). Nature of any permanent disability resulting from any accident/etc

Hospital admissions Works backwards. Age and why admitted Smoking Smoking habit of respondent and partner Drinking Alcohol consumption in last 7 days, other aspects of

drinking behaviour Diet Frequency of consumption of types of food, vegetarian or

other special diets Exercise Exercise at work and in daily life Height and weight Self-reported height and weight and assessment of weight (Based on work undertaken by:

• Barbara Jefferis, Institute of Child Health • Scott Montgomery, Karolinska Sjukhuset, Stockholm, Sweden • Richard Rowe and Barbara Maughan, Institute of Psychiatry • Samantha Parsons, Centre for Longitudinal Studies, Institute of Education, University

of London) Whilst reviewing the data on health, those named above identified a number of problems and produced a number of very valuable revised and derived variables. Details are reported below for each area of the areas identified above. Additional information on health was included in the self-completion and further information is given on these data are given in the relevant section below. General Health (Based on work undertaken by:

• Barbara Jefferis, Institute of Child Health • Richard Rowe and Barbara Maughan, Institute of Psychiatry)

Health Generally 1. How is your health generally? (HLTHGEN) -Rates similar in BCS70 and NCDS. For NCDS: Reported as Excellent (31%men, 29%women), good (51%men, 52%women), fair (15%men, 14%women), poor (3%men, 4%women), 2. How is your health over the last 12 months? (HLTHYR) - Rates in poorer categories perhaps a little lower in BCS70 as may be expected from their younger age


For NCDS, as in the NCDS4 (age 23) and NCDS5 (age 33) surveys, more men than women report excellent health - Excellent (29%men, 26%women), good (47%men, 46%women), fairly good (16%men, 16%women), not so good (8%men, 12%women). Overall, 76% of NCDS men and 72% NCDS women report good or excellent health in the last 12 months, more women report poorer health. This follows the same patterns as in the NCDS4 and NCDS5 year surveys. Specific Health Conditions (Based on work undertaken by:

• Barbara Jefferis, Institute of Child Health • Scott Montgomery,?? • Richard Rowe and Barbara Maughan, Institute of Psychiatry • Samantha Parsons, Centre for Longitudinal Studies, Institute of Education, University

of London) Variables with no valid cases The following variables are redundant, as they have no valid cases: OTHSKIN SKINCOND SL1AGE, SL112M, SL1DOC SKINCON7 EATING4 HERNIWH4 GYNAEP09 GYNAEP10 KINDDIA3 (CANCTY04 TO CANCTY12) UGICOCR5 Eating problems 1. The "Some other eating problem" category is the most common positive response but it is not

clear how this may be unpacked further, as EATOTH does not appear in the data set.

EATOTH is an open question which along with the answers to other open health questions, has recently been coded by CLS. These data will be made available shortly, along with the text of all open questions.

2, EATING4 variable included but all cases are missing (none in NCDS either). 3. Eating problems – some distributions and derived variables.



valid n

N % N % n % eatprob 66 0.3 6 21 0.1 22587 729 (3.2%) CMs reported an eating disorder. 729 asked name of eating problem(s) they have had in multi-coded variables eating1- eating4 (eating4 holds no information)

21951 96.8 1 728

el1age 21951 96.8 ^5 0.7 724 el112m 21951 96.8 1 0.1 1 0.1 727 349 (47.9%) CMs had an eating disorder in last 12 months. 349 asked if they had seen a doctor re: eating disorder el1doc 22331 349

Source: Samantha Parsons, Centre for Longitudinal Studies, Institute of Education, University of London.


Derived EATING PROBLEMS variables



Valid n

N % n % n % Eatprnow 95 0.4 22585 3 category variable combining information given in eatprob and el112m. Values: 0’no eating problems’ 1’previous eating problem’ 2’current eating problem’ bulemia 93 0.4 22587 Anorexia 93 22587 Swallow 93 22587 Otheatpr 93 22587 2 category variables. If CM gave valid answer to eatprob, used information in multi-coded eating1-eating4 variables. Values: 0’not this eating problem’ 1’yes, this eating problem

Source: Samantha Parsons, Centre for Longitudinal Studies, Institute of Education, University of London. Further details of these derived variables are to be found in the detailed documentation supplied by Samantha Parsons, which is included as Appendix 5. Blood Pressure For NCDS, 10% (n=592) men and 12.5% (n=708) of women report ever having had high blood pressure (perhaps gestational blood pressure may explain the difference?). This compares with 256 men and 356 women reporting ever had high blood pressure at 33 years. Approximately 6% of the cohort have had high blood pressure in the last 12 months. Cancers For NCDS, more cancers are reported in women than in men: 3.5 % compared to 2%. At age 41, 61 men and 202 women have had cancers, compared to 33 men and 130 women at age 33. Women report mostly breast, cervix and skin cancers, for men the most common cancers are skin and testes. 11 men and 38 women have had cancer in the past 12 months, all but 3 have seen their doctors about it. Diabetes For NCDS, approximately 1.7% of the cohort report ever having had diabetes, 98 men and 99 women, this compares with 39 men and 32 women reporting ‘ever had’ diabetes at age 33. 1.2% have had it in the past 12 months, most consulting their doctors about it in this time period. Allergy and headache For NCDS, 14% men and 15% women have had hay fever in the last 12 months (compared to 16% at age 33), 3% men and 5% women have seen a doctor about it in the last 12 months. Average age of onset for men was 18 years and for women, 19 years. Bronchitis For NCDS, 2% men and 3% of women in the sample have had bronchitis in last 12 months, with slightly fewer consulting their doctor about it in the last 12 months. Average age at which men first had bronchitis was 18 for men and 21 for women. Similar incidence rates are reported in the 33 year sample. Asthma For NCDS, 5% and 7% women have had asthma in the last 12 months. Average age men first had asthma was 17 years, and 24 years for women. 2% men and 5% women have had allergic rhinitis in last 12 months. Average age at which cohort first had allergic rhinitis was 24 years.


20% (n=1146) men and 25% (n=1444) women report ever having skin problems at age 41, in comparison at 33 15% men (n=820) and 20% women (n=1114) report ever having eczema or skin problems. Eczema or other skin problems For NCDS, 932 men and 632 women have had eczema or other skin problems (mostly psoriasis, acne and contact dermatitis, mouth ulcers, fungal infections) in the last 12 months. This is higher than the reported 12 month incidence at 33 years, for men 10% (n=556) and 14% (n=798 for women). Average age men and women first had eczema was around 20 years. Average age men first had acne was 14 years, for women it was 17 years. For psoriasis, it was 23.5 years for men and 22 years for women. The average age men and women first had cold sores was 21. 7% men and 19% of women have had severe headaches in the last 12 months. 2% men and 8% of women have seen a doctor about these in the last 12 months. Average age that cohort members first had severe headaches was 22. ME, back pain, eating problems, hernia and fits Less than 1% of the NCDS cohort report having suffered ME in the last 12 months. For NCDS, 22% of men and women have suffered persistent back pain, lumbago or sciatica. This compares with 51% men and 43% women reporting ever having lower back pain at age 33. 16% have had persistent back pain over the last 12 months, with 9% visiting their doctor about it. For NCDS, 2%(n=97) men and 4% (n=225) women have ever had eating problems. 7% men and 3% women report having had hernias. 1.6% (88) men and 1% (64) women have had a hernia in the past 12 months. For NCDS, 2% of the sample have ever had fits or convulsions, less than 1% in the past month. Average age men first had fits was 20, and women, 17 years. Women’s health For NCDS, 26% women reported problems with their periods- mostly heavy or painful. 15% had problems with their periods over the last 12 months, and 11% had seen their doctor about it. This compares with 18% reporting ever having persistent trouble with periods at age 33 and 14 % having persistent troubles over the last 12 months. At age 41, 22% had other gynaecological problems, commonly ovarian cysts, endometriosis and fibroids. 6% had gynaecological problems over the last 12 months and 5% had seen their doctor about it. Average age women first had gynaecological problems was 31. At age 33, 16% women reported ever havinh other gynecological problems and 10% in the last 12 months. 9% of women are currently taking the pill and 79 % have ever taken the pill. 26% men and 23% women have been sterilised or had vasectomy/ hysterectomy. Kidney, bladder problems, etc For NCDS: • 6% men and 10% women have ever had kidney or bladder problems (similar to age 33, 4% and

9% for men and women). In the 41 year old men it was mostly kidney or bladder stones, in women, mostly kidney or bladder infections.

• 2.2% man and 4.1% men have had kidney or bladder problems in last 12 months, most have

seen their doctor about it. • 5% women and 12% men have had irritable bowel syndrome.


• 5%men and 3% women have had duodenal or peptic ulcers. <1% men or women have had ulcers

in the last 12 months. • 4% men and 8% women have had gall stones in the last 12 months. • 0.3% of cohort had IBS in last 12 months and 0.2% had ulcerative colitis in last 12 months. A summary of the more prevalent conditions in the NCDS cohort is given below.



Most prevalent health conditions in the NCDS cohort

Ever had Men (%) Women (%)

In the last 12 months Men (%)

Women (%)

Accidents 45 25 Wheezing/ whistling chest

28 28 Back pain/ lumbago/ sciatica

22 22

Problems with periods 26 Persistent back pain 16 Painful periods 15 Wheezing/ whistling chest 19 Heavy periods 12 Skin problems 20 Gynecological problems

22 Migrane / headache 7 18

Ovarian cyst 4 Fibroids 3 Problems with periods 15 Endometriosis 3 Respiratory problems Hay fever 20 21 Winter cough (am) 14 13 Migrane / headache 13 27 Winter cough (am or pm) 13 14 Problems with sight or eyes

20 20 All year cough 10 10

Both eyes 15 16 Winter phlegm (am) 11 7 One eye 5 4 Winter phlegm (am or

pm) 8 6

Eczema 10 15 All year phlegm 8 5 High blood pressure 10 12.5 Asthma 10 12 Gynecological problems 6 Low/depressed/sad 8 15 High blood pressure 6 6 Bronchitis 8 13 Hay fever 6 6 Hearing problems 9 8 Gall stones 4 8 IBS 5 12 Bronchitis 7 3 Kidney / bladder problems

6 10 Asthma 5 5

Duodenal or peptic ulcers

6 3 Eating problems 2 4

Anxious/ jittery 3.5 6.5 Diabetes 1.2 Psoriasis 4 4 Allergic rhinitis 0.8 2 Phobic 2 4 ME <1 Rhinitis 3 6 Duodenal or peptic ulcers <1 Cold sores 2 4 IBS 0.3 Cancers 2 3.5 Ulcerative colitis 0.2 Acne 3 2 Fungus infections 3 2 Contact dermatitis 2 3 Gallstones 0.8 4 Diabetes 1.7 Fits/ convulsions <2 Mouth ulcers 0.8 1

Source: Barbara Jefferis, Institute of Child Health


Long-term health conditions The following variables are redundant, as they have no valid cases: LSIIMWK4 LSIIMWK5 LSIIMWK6 LSIIMWK7 LSIIMWK8 LSIIMWK9 LSIAGE4 LSIAGE5 LSIAGE6 LSIAGE7 LSIAGE8 LSIAGE9 Respiratory problems For NCDS: • 14% men and 13% women cough first thing in the morning in winter. • 11% men and 7% women bring up phlegm from chest in the morning in winter. • 8% men and 6% women usually bring up phlegm from chest during day or night in winter. • 12.6% men and 14.2% women usually cough day or night in winter. • 10% men and women cough on most days for at least 3 months of the year. • 8% men and 5% bring up phlegm from chest most days for at least 3 months of the year. • These responses to the respiratory health questions are very similar at age 33. • 2.3% men and 1.9% women cannot walk. • 6% men and 11% women are ever out of breath walking on level ground (more than at 33 years). • 28% men and women have ever had whistling or wheezing chest. 18.5% have had this in the last

12 months. Of these, 60% have had it less than 5 times and 40% five times or more. • 8% men and 11% women in the cohort have used an inhaler or other prescribed asthma medicine

in the last 12 months. Mental health Whilst reviewing the data on health, Richard Rowe and Barbara Maughan made a number of observations about the quality and utility of the data. These are summarised below. 1. MHPROBS1 a) MHPROBS1 contain 98 and 99 codes; these need labelling.

As noted above, the following missing values will be found within the data:




b) It would be much easier to code MHPROBS1-9 by disorder as the age of onset variables are coded. ie. MHPROBS1 would be a binary variable indicating the presence or absence of depression etc. As noted above, variable names have been allocated through CAPI (Blaise). MHPROBS is the name of the CAPI variable and where the question has been repeated, Blaise places adds numerical suffix strating with 2, hence the sequence fro MHPROBS (and all repeated variables) is MHPROBS, MHPROBS2,…,MHPROBSn.

c) Prevalence by sex looks plausible (see table below). 2. Age of onset variables a) Depression age at onset: MHAGE more consistently called MHAGE1. Without the number, it

looks like this may be a summary of the MHSPECS that include a numerical identifier. 99 codes need to be labelled as missing.

See comments above.

b) All look consistent with mhprobs variables. 3. MHSPEC a) Similarly to age of onset, mhspec should be called mhspec1. b) MHSPEC labelled ‘seen specialist/been to hosp’ whereas mhprobs say ‘seen specialist’. It

would be helpful to make this consistent. c) A lot of missing data on these variables. Of those who report depression since refdate. 52%

are missing on mhspec. Similar levels of missing regarding other disorders. It is not clear why this a problem here, particularly as it is not a problem for the mhstill variables.

4. MHSTILL a) Similarly to age of onset and MHSPEC, MHSTILL should be called MHSTILL1. b) Rates look plausible (see table below)



Prevalence of mental health problems reported in NCDS/BCS70

In entire sample

In those having experienced problem since refdate (males and females

combined) Disorder

Male

Female

Still has problems most of the time

Still has problems

occasionally

No longer has

problems NCDS

14%

(N=811)

29%

(N=1655)

14%

41%

45%

Depression

BCS-70

13%

(N=702)

28%

(N=1631)

10%

38%

52%

NCDS

6%

(N=306)

10%

(N=548)

18%

48%

33%

Anxiety

BCS-70

5%

(N=283)

8%

(N=447)

17%

45%

38%

NCDS

3%

(N=165)

6%

(N=332)

20%

49%

30%

Phobias

BCS-70

4%

(N=196)

6%

(N=317)

18%

43%

39%

NCDS

0.4%

(N=22)

0.4%

(N=24)

23%

34%

41%

Overexcited/over confident

BCS-70

0.5% (N=26)

0.4%

(N=21)

25%

47%

28%

NCDS

1%

(N=54)

1%

(N=60)

33%

41%

25%

Compelled to repeat activities

BCS-70

1% (N=54)

1%

(N=62)

33%

40%

28%

Hallucinations

NCDS

1%

(N=40)

1%

(N=40)

24%

48%

29%


BCS-70

1%

(N=50)

1%

(N=49)

24%

34%

41%

NCDS

2%

(N=93)

1%

(N=64)

26%

25%

49%

Alcohol problems

BCS-70

2% (N=135)

0.6%

(N=37)

21%

17%

62%

NCDS

0.6%

(N=33)

0.3%

(N=18)

35%

20%

45%

Drug problems

BCS-70

2%

(N=114)

0.7%

(N=41)

21%

14%

65%

NCDS

0.04% (N=2)

0.9%

(N=51)

-

-

-

Bulimia

BCS-70

0.2% (N=9)

1.7%

(N=98)

-

-

-

NCDS

0.1% (N=4)

1.3%

(N=73)

-

-

-

Anorexia

BCS-70

0.1% (N=5)

2%

(N=111)

-

-

-

NCDS

0.2% (N=9)

0.2% (N=9)

-

-

-

Swallowing problems

BCS-70

0.1% (N=4)

0.1% (N=8)

-

-

-


Seeing and hearing For NCDS: • 20% of the cohort report problems with sight or eyes: 15% in both eyes and 5% in one eye. • 9% have hearing problems: evenly distributed between problems with one ear or with both. • 8% of the cohort are short sighted, 4% long sighted and 2.5% have other sight or eye conditions. • 60% men and 40% women do not wear glasses or contact lenses. 35% men wear glasses and

5% wear contact lenses- women wear lenses more often than men do. 5% men and 7% women wear both contact lenses and glasses.

• 65% men and 58% women read books, watch television and see into the distance without

problems. • 90% of those reporting their everyday vision thought it was average or better than average and

7% reported slightly abnormal vision. • 5% men and 3% women have some hearing problems (tinnitus, repeated ear infections and

discharge) • <0.5% wear hearing aids.


Other conditions 1. The following variables are redundant, as they have no valid cases. HOSUP33 HOSUP36 HOSUP38 HOSUP39 HOSUP41 HOSUP42 HOSUP43 HOSUP44 HOSUP45 HOMORE14 HOAGE15 HOMORE15 HOAGE16 HOSUP46 HOSUP47 HOSUP48 HOMORE16 HOAGE17 HOSUP49 HOSUP50 HOSUP51 HOMORE17 HOAGE18 HOSUP52 HOSUP53 HOSUP54 HOMORE18 HOAGE19 HOSUP55 HOSUP56 HOSUP57 HOMORE19 HOAGE20 HOSUP58 HOSUP59 HOSUP60 HOMORE20 2. For NCDS, 5% men and 7% women have another health condition. This is mostly monitored, either by hospital or clinic or by GP. 20% men and 105 women report that their conditions are not monitored. 0.2% men and 0.5% women are monitored for more than one other health condition.



Accidents/injuries 1. The following variables are redundant, as they have no valid cases: ACCIDAN8 ACCIDAN9 ACCINJ18ACCINJ24 ACCINJ30 ACCINJ36 2. 45% men have had accidents since “refdate”, this appears to be age 33. Of these, 13% have had accidents at work, 10% a sports accident and 8% a road accident. Of the 25% women cohort members who have had accidents, 8% have had road accidents, 7% have had an accident in the home and 4% have had an accident at work. 3. For NCDS, men and women differ in the number of accidents that they have seen a doctor for. 13% men have had a sports accident and 12% have had an accident at work. 6% have had an accident at home and 5% have been a driver or passenger in a road accident. 7% women have had an accident at home and 6% have been a driver or a passenger in a road accident. 4% have had an accident at work or an ‘other’ type of accident. 3% have had a sports accident. 2% have had a violent assault or mugging compared to 4% men. 4. Further details of aspects of the information about accidents, and a number of derived variables are given below.



valid n

N % N % N % Multi-coded questions ask if CM had an accident, and if yes, what type of accident. Accidan1 accidan2 accidan3 accidan4 accidan5 accidan6 accidan7 accidan8 accidan9

66 0.3 8 0 24 0.1 22582

12458 (55%) reported no accidents with 10124 (45%) reporting they had some kind of accident. These CMs asked how many accidents they had had in total, the age and type of their most recent accident Accidno 12524 55.2 ^10 0.1 ^42 0.4 10104 Accage 12576 55.4 ^1 0 ^18 0.2 10085 Accwhy 12576 55.4 1 0 2 0 10101

Source: Samantha Parsons, Centre for Longitudinal Studies, Institute of Education, University of London. Derived ACCIDENT variables



valid n

N % N % N % 2-category derived variables from multi-coded accidan1-accidan9. Each variable gives % experiencing type of accident: road accident (pedestrian), road accident (driver/passenger), at work, at home, at school/college, playing sport, other type, violent assault, sexual assault. Values: 0’not experienced’ 1’yes, experienced’. accroadp accroadd accwork acchome accschcl accsport accother vassault asssex

98 0.4 22582


Source: Samantha Parsons, Centre for Longitudinal Studies, Institute of Education, University of London. Further details of these derived variables are to be found in the detailed documentation supplied by Samantha Parsons, which is included as Appendix 5.


Smoking 1. Missing values: There are occasional 8s and 9s on SMOKING, NOFCIGS, AGEQUIT, OTHSMOKE and PARTCIGS

As noted above, the following missing values will be found within the data:



2. Basic distributions:

Men

Women

BCS70

(n=5438) %

NCDS

(n=5600) %

BCS70

(n=5767) %

NCDS

(n=5775) %

Never smoked cigarettes

41.8

43.7

46.2

45.5

Used to smoke, don’t at all now

18.2

26.0*

19.7

24.5*

Smoke occasionally

8.1

4.6

7.3

4.0

Smoke every day

31.9

25.7

26.8**

26.0**

* Mean age of quitting is similar in men and in women, at around 30.5 years. ** Of the current smokers, consumption in men and women peaks at 20/day (31%) with only 5% smoking more than 30 /day. The other 2 peaks are at 10 and 15 per day. Average number of cigarettes smoked per day is 23 for men and 18 for women.


3. Further details of aspects of the information about smoking, and a derived variable giving categories for number of cigarettes smoked are given below.

Variable system-missing user-missing: 8 (^ indicates 98;

^îndicates 998)

user-missing: 9 (^ indicates 99;

^îndicates 999)

valid n

n % n % N % Smoking 66 0.3 8 26 0.1 22580 6222 (27.4%) CMs smoke every day. 6222 sked how many they smoke a day Nofcigs 16458 72.6 ^^16 0.3 6206 6347 (28%) CMs ex-smokers or occasional smokers. 6347 asked if ever smoked cigarettes regularly exsmoker 16333 72.0 6347 4532 (71.4%) CMs had smoked regularly. 4532 asked age last smoked regularly Agequit 18148 80.0 ^6 0.1 4526 othsmoke 2382 10.5 8 29 0.1 20261 5864 (25.9%) live with someone who smokes in CM home. 5864 asked whether this was spouse/partner or someone else whosmoke 16816 74.1 5864 4521 (77.1%) CMs said partner smoked. 4521 asked how many spouse/partner smoked a day Partcigs 18159 80.1 1 27 0.1 4493

Source: Samantha Parsons, Centre for Longitudinal Studies, Institute of Education, University of London. Derived SMOKING variable

variable system-missing user-missing: 8 (^ indicates 98;

^îndicates 998)


^îndicates 999)

valid n

N % n % n % Smoke 116 0.5 22564 6 category variables combining information from smoking and nofcigs. Values:0’never smoked’ 1’ex-smoker’ 2’occasional smoker’ 3’up to 10 a day’ 4’11 to 20 a day’ 5’more than 20 a day’.

Source: Samantha Parsons, Centre for Longitudinal Studies, Institute of Education, University of London. Further details of these derived variables are to be found in the detailed documentation supplied by Samantha Parsons, which is included as Appendix 5.



Drinking 1. Further details of aspects of the information about alcohol consumption, and a number of derived variables are given below.


^îndicates 998)


^îndicates 999)

valid n

n % n % n % Drinks 66 0.3 7 27 0.1 22580 18260 (80.7%) CMs drank alcohol regularly (daily, weekly or monthly). 18260 asked how many of each type of alcoholic drink they had in last 7 days prior to interview beer 4420 19.5 ^^^16 0.1 18244 Spirits 4420 19.5 ^^^1 ^^^16 0.1 18243 Wine 4420 19.5 ^^^1 ^^^3 18256 Sherry 4420 19.5 ^^^1 ^^^1 18258 Pops 4420 19.5 ^^^1 ^^^1 18258 othdrink 4420 19.5 ^^^1 18259 cage1 493 2.2 5 22182 5521 (24.9%) CMs ever felt should cut down on their drinking. 5521 asked if felt this in last year cage2 17159 75.7 1 5520 cage3 493 2.2 22187 2193 (9.7%) CMs ever felt criticised about their drinking. 2193 asked if felt this in last year cage4 20487 90.3 2193 cage5 493 2.2 22187 2693 (11.9%) CMs ever felt bad or guilty about their drinking. 2693 asked if felt this in last year cage6 19987 88.1 2693 cage7 493 2.2 1 1 22185 874 (3.9%) CMs ever had a drink first thing in A.M. to steady hands. 874 asked if they had in last year cage8 21806 96.1 874 cage9 4208 18.6 4 0 18468 2539 (11.2%) CMs ever drink during breaks at work. 2539 asked how often they did this cage10 20141 88.8 2 2537

Source: Samantha Parsons, Centre for Longitudinal Studies, Institute of Education, University of London. Derived DRINKING variables


^îndicates 998)


^îndicates 999)

valid n

n % n % n % drkunits 44250 18230 Continuous variable giving total number of alcohol units CM had in 7 days prior to interview. Combines information given in beer, spirits, wine, sherry and pops. Range from 0-427. Mean: 19.26, Median: 11.00, sd: 25.77 Drkcut 499 2.2 22181 3 category variable combining cag1 and cage2. Values: 0'no' 1'yes, previously' 2'yes, last 12 months'. Drkcrit 493 2.2 22187 3 category variable combining cag3 and cage4. Values: 0'no' 1'yes, previously' 2'yes, last 12 months'. Drkbad 493 2.2 22187 3 category variable combining cag5 and cage6. Values: 0'no' 1'yes, previously' 2'yes, last 12 months'. drkhand 495 2.2 22185 3 category variable combining cag7 and cage8. Values: 0'no' 1'yes, previously' 2'yes, last 12 months'. drkwork 4214 18.6 18466 3 category variable combining cag9 and cage10. Values: 1'yes, special occasions' 2'yes, 2 or 3 times a month' 3'yes, 2 or 3 times a week' 4'yes, most days'.



Further details of these derived variables are to be found in the detailed documentation supplied by Samantha Parsons, which is included as Appendix 5. 2. The information gathered on drinking also includes 10 items from the CAGE scale, see below.

Cage variables Variable Label CAGE1 Ever felt ought to cut down on drinking CAGE2 Felt ought to cut down on drinking in last year? CAGE3 Have people annoyed you by criticising your drinking CAGE4 People annoyed CM re: drinking in last year CAGE5 Ever felt bad or guilty about your drinking CAGE6 Felt bad/guilty about drinking in past year? CAGE7 Ever had a drink first thing in am to steady hands etc CAGE8 Had drink first thing in am in past year? CAGE9 Ever have an alcoholic drink during breaks in daily work CAGE10 Freq of drinking during work breaks in past year Items 1, 3, 5, 7 may be used to create a drinking problem scale; scores of 2 or more positively endorsed items on this scale have been shown to provide a good indicator of alcohol problems in community samples (King, M. (1986). At risk drinking among general practice attenders. Validation of the CAGE questionnaire. Psychological Medicine 16, 213-217; Mayfield, D., McCloed, G. & Hall, P. (1984). The CAGE questionnaire. Validation of a new alcohol screening instrument. American Journal of Psychiatry 131, 1121-1123). SPSS syntax to generate CAGETOT (total score) and CAGEDI (low vs high) is given below:. *CAGE total score (count of positively endorsed items) and binary alcohol problem indicator (see Mayfield et al, 1984; King, 1986 for details). *set no to 0 and yes to 1 and declare missing values. recode cage1 cage3 cage5 cage7 (2 = 0). mis val cage1 cage3 cage5 cage7 (8, 9). val lab cage1 cage3 cage5 cage7 0 'no' 1 'yes'. *compute CAGE total score. compute cagetot = cage1 + cage3 + cage5 + cage7. recode cagetot (sysmis = 9). var lab cagetot 'Cage (ever), total score'. val lab cagetot 9 'not known'. *compute binary indicator of alcohol problems. compute cagedi = cagetot. recode cagedi (2 through 4 = 1) (0, 1 = 0) (sysmis = 9). mis val cagetot cagedi (9). var lab cagedi 'Cage (any drinking problem ever), 2+ cutoff cagetot'. val lab cagedi 0 'no' 1 'yes' 9 'not known'.



Diet 1. For NCDS: • 45% men and 60% women eat fresh fruit once / day or more. 20% men and 14% women eat it

less than once a week or occasionally. • 11% men and 25% women eat salads or raw vegetables once a day or more. 25% men and • 15% women eat salads or raw vegetables less than once a week. • 35% men and 43% women eat cooked vegetables at least once per day. • 66% men and 53% women eat food fried in vegetable oil less than once a week. 20 % men and

10% women eat food fried in vegetable oil 3-6 days a week. • 8% men and 5% women eat food fried in hard fat 1-2 times a week. 60% men and 73% women

never eat food fried in hard fat. • 53% men and 35% women eat chips more than 1-2 days a week. • 50% men and 40% women eat eggs less frequently than twice a week. • 9% men and 21% women eat sweets and chocolates once a day or more. 25% men and women

eat sweets and chocolates only occasionally (<1 day /week). • 23% men and 21% women eat cakes or biscuits at least once per day. 23% men and 26%

women eat cakes and biscuits only occasionally. • 40% men and 47% women eat whole meal bread or rolls 3-6 days per week. 68% men and • 60% women eat other bread or rolls 3-6 times a week. • 32% men and 24% women eat red meat 3-6 days a week. • 34% men and 40% women eat poultry 3-6 days a week. • 70% men and women eat fresh fish less than one day a week. • 76% men and 70% women eat pulses less frequently than one day per week. • 2% men and 3% women are vegetarians. • 3% men and 7% women are on a special diet- sometimes prescribed by doctor.


2. Further details of aspects of the information about diet are given below.



valid n

N % N % n % Fruit 66 0.3 9 28 0.1 22577 Eggs 66 0.3 10 29 0.1 22575 Salads 66 0.3 10 28 0.1 22576 cookdveg 66 0.3 10 26 0.1 22578 Oilfried 66 0.3 11 37 0.2 22566 Chops 66 0.3 11 45 0.2 22558 Sweets 66 0.3 11 27 0.1 22576 Cakes 66 0.3 11 26 0.1 22577 whlbread 66 0.3 11 27 0.1 22576 Othbread 66 0.3 11 26 0.1 22577 Redmeat 66 0.3 11 26 0.1 22577 Poultry 66 0.3 11 26 0.1 22577 Fish 66 0.3 11 26 0.1 22577 Pulses 66 0.3 11 27 0.1 22576 Veggy 66 0.3 9 26 0.1 22579 808 (3.6%) CMs vegetarian. 808 asked the type of vegetarian they were Vegtype 808 Spshdiet 66 0.3 9 26 0.1 22579 1061 (4.7%) CMs followed a special diet. 1061 asked the type of special diet they followed. Only 1009 (95% of 1061) asked if the diet had been recommended by a doctor Diettype 1061 Dietdoc 1009 Exercise 66 0.3 10 27 0.1 22577 17263 (76.3%) CMs do regular exercise. 17263 asked how often they take regular exercise and if they become breathless/sweaty during exercis Breathle 2 0 17261 Sweat 3 0 17260




Exercise For NCDS: • 75% men and 72% women do regular exercise. • 15% men and 20 % women exercise every day. • 61% men and 67% women do regular exercise 2-3 days a week. • 38% men and 23% women report getting out of breath or sweaty most times that they exercise.



CITIZENSHIP AND VALUES Citizenship and Values Involvement with organisations, voting behaviour and

intentions, political alignment, trade union membership, religion, newspaper readership, car ownership, values, political activity

Review of these data by the CLS NCDS/BCS70 Team and by advisors has, to date, revealed no particular problems, nor have any derived variables been developed. The data will continue to be investigated and, if necessary, detailed updates will be made available, and the dataset will be updated.



SELF-COMPLETION

SELF-COMPLETION Your views Attitude statements How you get on with your husband, wife or partner

Includes Locke-Wallace

Some more of your views Attitude statements How you feel Malaise Inventory Your skills How good at skill/is skill used at work More of your views Attitude statements How you feel about your life so far GHQ 12 More of your views Attitude statements School exclusion and truancy Number of temporary/permanent suspensions/exclusions;

frequency of truancy Contact with the police and crime Number of times moved on, questioned, warned, taken to

police station, cautioned, found guilty by a court Use of illegal drugs Whether tried number of specific drugs ever/in last 12

months Your views/Some more of your views/More of your views (Based on work undertaken by John Preston, Centre for Longitudinal Studies, Institute of Education, University of London) In reviewing the data for the questions on ‘Views’ in the self-completion (CASI), John Preston developed the following ten attitude scales:

Derived Alpha coefficients Scale variable NCDS BCS70 1 Anti-racism TOLERACE 0.8000 0.8222 2 Left-Right LEFTR 0.7313 0.6792 2 Support for Authority AUTHORIT 0.6470 0.6176 4 Support for traditional marital values PROFAM 0.6620 0.6345 5 Effect of children on quality of life NOKIDS 0.6141 0.6232 6 Support for working Mothers MUMWORK 0.6880 0.6782 7 Political cynicism POLCYN 0.6742 0.6526 8 Technophobia ANTITECH 0.5892 0.5732 9 Environmentalism ENVIRON 0.4664 0.5070 10 Support for the work ethic WORKETH 0.5388 0.5469

The scales were constructed using the 50 ‘Views’ items of the self-completion questionnaire (see below), although a small number were not included in the scales, and did not appear to form any clear factors of their own. Data for NCDS and BCS70 was analysed separately, although the factors identified in each were, fortuitously, identical. Principal components analysis was employed using varimax rotation to orthogonal simple structure. The number of factors chosen was based upon those with Eigen values greater than one, followed by the ‘scree’ test to determine that point at which Eigen values leveled off. Factors were examined for interpretability and to ensure that at least three items had loadings on each of them. Cronbach’s alpha was employed as a test of internal consistency of the factors. In all cases, alpha scores were above 0.5 which indicates that it is acceptable to consider the items as forming an identifiable factor. For each factor, the derived variable identified above is the mean of the scores for each item within the component. Further details of the development of these derived variables and appropriate SPSS


code are to be found in the detailed documentation supplied by John Preston, which is included as Appendix 8.

Questions on ‘Views’ in the self-completion (CASI) Variable Label VariableLabel LR1 (SC) Big business benefits owners at expense of workers LR5 (SC) Ordinary people dont get fair share of nations wealth AR1 (SC) Mixed race marriage is OK AR4 (SC) Wouldnt mind working with people from other races E1 (SC)Problems in the environment not that serious IT2 (SC) Computers enrich the lives of users A1 (SC) The law should be obeyed even if wrong MOR4 (SC) Marriage is for life? C1 (SC) Unless have kids will be lonely in old age A4 (SC) Give law breakers stiffer sentences PC1 (SC) No political party would benefit me E3 (SC) The environment vs economic growth L1* (SC) More likely to get better job if do training/educ WE2 (SC) If I didn t like a job Id pack it in MOR1 (SC) Divorce is too easy to get these days LR6 (SC) Government should redistribute income LR2 (SC) Private schools should be abolished WM4 (SC)A mother and family happier if she goes out to work WM1 (SC) Pre-school kids suffer if mum works L3 (SC) Learning new things boosts confidence IT1 (SC) Computers at work destroying peoples skills C3* (SC) Having children interferes with parents freedom WE1 (SC) Any job is better than being unemployed A5 (SC) Young people dont have respect for trad values MOR2 (SC) Married people happier than unmarried WM3 (SC) Kids benefit if mum has job outside home? A2 (SC) Death penalty for some crimes LR7 (SC) One law for the rich and one for poor? MOR3 (SC) Couples with kids should not separate MOR5 (SC) Women should have the right to an abortion? LR3 (SC) Management get the better of employees PC3 (SC) Politicians in politics for own benefit? AR2 (SC) Wouldnt mind if family of diff race moved next door L4* (SC) Effort of getting quals more trouble than worth? C2 (SC) Can have fulfilling life with no kids AR5 (SC) Not want another race person as my boss? L2* (SC) Knowing right people helps more than quals to get job MOR6 (SC) Alright for unmarried people to have kids? PC2 (SC) No difference which political party is in power in GB A6 (SC) Schools teach children to obey authority? LR4* (SC) Take out own private health care, stop relying on NHS IT4 (SC) Every family should have a computer? E2 (SC) Preserving environment most important WE3 (SC) Important to hang onto job even if unhappy? A3 (SC) Censorship is needed to uphold morals WM5 (SC) Dads job is to earn money;mums to stay home? WM2 (SC) Family life suffers if mum working ft C4 (SC) People with no kids are missing out? AR3 (SC) Would mind kids going to school with diff races IT5* (SC) Learning to use a computer more trouble than worth? * Variable not included in the scales and not appearing to form any clear factors of their own.



How you feel (Malaise Inventory) (Based on work undertaken by Richard Rowe & Barbara Maughan, Institute of Psychiatry) The Malaise Inventory provides a measure of for assessing psychiatric morbidity, developed by the Rutter and others at the Institute of Psychiatry from the Cornell Medical Index (Rutter M, Tizard J, and Whitemore K (1970) Education, Health and Behaviour. London). It is a 24-item self-completion scale and has been included in earlier NCDS and BCS70 follow-ups. The 24 variables included on the dataset are identified below.

Variable Label MAL01 (SC) Do you often have backache? MAL02 (SC) Do you feel tired most of the time? MAL03 (SC) Do you often feel miserable or depressed? MAL04 (SC) Do you often have bad headaches? MAL05 (SC) Do you often get worried about things? MAL06 (SC) Usually have difficulty falling or staying asleep? MAL07 (SC) Usually wake unnecessarily early in morning? MAL08 (SC) Do you wear yourself out worrying about health? MAL09 (SC) Do you often get into a violent rage? MAL10 (SC) Do people often annoy and irritate you? MAL11 (SC) Have you had twitching of face/neck/shoulders? MAL12 (SC) Often suddenly become scared for no reason? MAL13 (SC) Often scared to be alone without friends near? MAL14 (SC) Are you easily upset or irritated? MAL15 (SC) Are you frightened of going out alone? MAL16 (SC) Are you constantly keyed up and jittery? MAL17 (SC) Do you suffer from indigestion? MAL18 (SC) Do you suffer from an upset stomach? MAL19 (SC) Is your appetite poor? MAL20 (SC) Does every little thing get on your nerves? MAL21 (SC) Does your heart often race like mad? MAL22 (SC) Often have bad pains in your eyes? MAL23 (SC) Troubled with rheumatism or fibrositis? MAL24 (SC) Have you ever had a nervous breakdown?

Typically, the Malaise inventory has been used to produce:

• Total score: The sum of the positive responses to the 24 items

• Binary categorisation: Scores 0-6 categorised as ‘Normal’ and 7 and higher as ‘Depressed’ Recent analysis has shown evidence that the responses to the items of the Malaise Inventory may represent two separate psychological and somatic sub-scales rather than a single underlying factor:

• Psychological subscale: 15 Items - 2, 3, 5, 6, 7, 8, 9, 10, 12, 13, 14, 15, 16, 19, 20

• Somatic subscale: 8 items - 1, 4, 11, 17, 18, 21, 22, 23


SPSS code to generate all four Malaise measures is given below. *Malaise Inventory – Total Score (items 1-24=1) compute maltot=0. do repeat x=mal01 to mal24. if(x eq 1)maltot=maltot+1. end repeat. count malmiss=mal01 to mal24(sysmis,8,9). compute malvalid=24-malmiss. if(malvalid ge 19)maltot=maltot*24/malvalid. if(malvalid lt 19)maltot=-1. missing values maltot(-1). *Malaise Inventory – Binary categorisation (1-6=1;7-24=2) compute malaise=maltot. recode malaise(0 thru 6=1)(6 thru 24=2). * Malaise Inventory – Psychological subscale (items 2, 3, 5, 6, 7, 8, 9, 10, 12, 13, 14, 15, 16, 19, 20=1) count malpsych=mal02 mal03 mal05 mal06 mal07 mal08 mal09 mal10 mal12 mal13 mal14 mal15 mal16 mal19 mal20 (1). * Malaise Inventory – Somatic subscale (items 1, 4, 11, 17, 18, 21, 22, 23=1) count malsom=mal01 mal04 mal11 mal17 mal18 mal21 mal22 mal23 (1). if missing(maltot)malpsych=-1. if missing(maltot)malsom=-1. missing values malpsych malsom (-1). var labels maltot "Total Malaise score (sum of 1s)"/ malaise "Malaise categories"/ malpsych "Malaise: Psychological subscale"/ malsom "Malaise: Somatic subscale". value labels malaise 1 'Normal' 2 'Depressed'. fre maltot malaise malpsych malsom.


For more information see:

Rodgers B., Pickles, A., Power, C., Collishaw, S. & Maughan, B. (1999). Validity of the Malaise Inventory in general population samples. Social Psychiatry and Psychiatric Epidemiology 34, 333-341)

Available at: http://link.springer.de/link/service/journals/00127/bibs/9034006/90340333.htm


Your skills (Based on work undertaken by Samantha Parsons, Centre for Longitudinal Studies, Institute of Education, University of London) After reviewing the self-completion data, Samantha Parsons produced the following information about responses to questions included on work-related skills Further details of all variables are to be found in the detailed documentation supplied by Samantha Parsons, which is included as Appendix 5.



valid n

n % n % n % SKILL1A 283 1.2 3 5 22344 CM reported ‘good’ ‘okay’ or ‘poor’ communicating skills. 21544 (96%) reported if they used this skill at work. SKILL1B 1136 5.0 1 21543 SKILL2A 283 1.2 3 4 22268 CM reported ‘good’ ‘okay’ or ‘poor’ number/calculation skills. 21477 (96%) reported if they used this skill at work. SKILL2B 2103 5.3 1 21476 SKILL3A 283 1.2 3 7 19726 CM reported ‘good’ ‘okay’ or ‘poor’ computer/IT skills. 19173 (86%) reported if they used this skill at work. SKILL3B 3507 15.5 1 19172 SKILL4A 283 1.2 3 7 22219 CM reported ‘good’ ‘okay’ or ‘poor’ team working skills. 21469 (96%) reported if they used this skill at work. SKILL4B 1211 5.3 3 21466 SKILL5A 283 1.2 3 6 22297 CM reported if they were ‘good’ ‘okay’ or ‘poor’ at learning new skills. 21517 (96%) reported if they used this skill at work. SKILL5B 1163 5.1 1 21516 SKILL6A 283 1.2 3 7 22308 CM reported ‘good’ ‘okay’ or ‘poor’ problem solving skills. 21515 (96%) reported if they used this skill at work. SKILL6B 1165 5.1 1 21514 SKILL7A 283 1.2 3 6 21909 CM reported ‘good’ ‘okay’ or ‘poor’ tool use skills. 21149 (94%) reported if they used this skill at work. SKILL7B 1531 6.8 1 21148 SKILL8A 283 1.2 3 19 0.1% 20975 CM reported ‘good’ ‘okay’ or ‘poor’ caring skills. 20217 (90%) reported if they used this skill at work. SKILL8B 2463 10.9 1 21216 SKILL9A 283 1.2 3 5 21228 CM reported ‘good’ ‘okay’ or ‘poor’ finance/accounting skills. 20523 (92%) reported if they used this skill at work. SKILL9B 2157 9.5 1 20522


Queries about problems with these variables should be directed to the User Support Group

([email protected])


How you feel about your life so far (General Health Questionnaire) The General Health Questionnaire (GHQ) is a self-administered screening test, designed to identify short-term changes in mental health (depression, anxiety, social dysfunction and somatic symptoms). It is a pure state measure, responding to how much a subject feels that their present state "over the past few weeks" is unlike their usual state. It does not make clinical diagnoses and should not be used to measure long-standing attributes. There are four different versions:

• GHQ12 - A quick screener for survey use.

• GHQ28 - Used to examine a profile of scores.

• GHQ30 - A screener with 'physical' element items removed.

• GHQ60 - Used to identify cases for more intensive examination. The GHQ12 was included in the self-completion (CASI) as it is very quick to administer. Nonetheless, it is just as reliable, valid and sensitive as the other longer versions. The 12 variables included in the dataset are identified below.

Variable Label GHQ1 (SC)...can concentrate on what you are doing? GHQ2 (SC)...lost much sleep over worry? GHQ3 (SC)...felt you were playing a useful part in things? GHQ4 (SC)...felt capable of making decisions? GHQ5 (SC)...felt constantly under strain? GHQ6 (SC)...felt could not overcome difficulties? GHQ7 (SC)...been able to enjoy normal activities? GHQ8 (SC)...been able to face up to your problems? GHQ9 (SC)...been feeling unhappy and depressed? GHQ10 (SC)...been losing confidence in yourself? GHQ11 (SC)...been thinking yourself as worthless? GHQ12 (SC)...been feeling reasonable happy?

The GHQ12 is also quick to score. There are two approaches: • Binary scoring -This the usual way of scoring the GHQ when it is to be used for case

identification. The advantage of using this method of scoring is that by weighting the answer options 0, 0, 1, 1, it avoids problems due to "middle users". To score you simply assign 0 for the first two answer options, 1 for the second two answer options and total the scores. (If used as a screener, rather than for survey work, this score can be compared to a threshold).

• Likert scoring - The authors suggest that if subscale scores are required for GHQ-28, there are

marginal advantages in using the Likert method of scoring which assigns weights of 0, 1, 2 and 3 to each answer option. To score you simply assign 0 for the first answer option, 1 for the second answer option, 2 for the third answer option and 3 for the fourth answer option and total the scores. (If used as a screener, rather than for survey work, this score can also be compared to a threshold).


SPSS code to generate both Binary and Likert GHQ12 scores is given below. * GHQ12 Binary Score (weighting the answer options 0, 0, 1, 1) fre ghq1 ghq2 ghq3 ghq4 ghq5 ghq6 ghq7 ghq8 ghq9 ghq10 ghq11 ghq12. missing values ghq1 ghq2 ghq3 ghq4 ghq5 ghq6 ghq7 ghq8 ghq9 ghq10 ghq11 ghq12 (8,9). recode ghq1 ghq2 ghq3 ghq4 ghq5 ghq6 ghq7 ghq8 ghq9 ghq10 ghq11 ghq12 (1 thru 2=0) (3 thru 4=1) (else=copy) into ghq1bi ghq2bi ghq3bi ghq4bi ghq5bi ghq6bi ghq7bi ghq8bi ghq9bi ghq10bi ghq11bi ghq12bi. compute ghq12bi = ghq1bi+ghq2bi+ghq3bi+ghq4bi+ghq5bi+ghq6bi+ghq7bi+ghq8bi +ghq9bi+ghq10bi+ghq11bi+ghq12bi. execute. * GHQ12 Likert Score (weighting the answer options 0, 1, 2 and 3) recode ghq1 ghq2 ghq3 ghq4 ghq5 ghq6 ghq7 ghq8 ghq9 ghq10 ghq11 ghq12 (1=0) (2=1) (3=2) (4=3) (else=copy) into ghq1lik ghq2lik ghq3lik ghq4lik ghq5lik ghq6lik ghq7lik ghq8lik ghq9lik ghq10lik ghq11lik ghq12lik. compute ghq12lik = ghq1lik+ghq2lik+ghq3lik+ghq4lik+ghq5lik+ghq6lik+ghq7lik +ghq8lik+ghq9lik+ghq10lik+ghq11lik+ghq12lik. var labels ghq12bi “GHQ12 score”/ ghq12lik “GHQ12 Likert score”. fre ghq12bi ghq12lik.


A useful website about the General Health Questionnaire including the scoring methods is:

www.nfer-nelson.co.uk/html/health/products/ghq.htm


Contact with the police and crime (Based on work undertaken by Richard Rowe & Barbara Maughan, Institute of Psychiatry) Whilst reviewing the data from the self-completion, Richard Rowe & Barbara Maughan made a number of observations about the quality and utility of the data. These are summarised below. COURT, POLICE1 TO POLICE5:

• As expected more men, and more respondents in BCS70, report police contacts or being found guilty by a court (e.g. 23.5% of men in BCS70 have been found guilty by a court).

• 64.8% of men in BCS70 report being stopped and questioned. Are such high rates plausible?

• NB: 170 people who have been found guilty in a court have not reported any police contacts.

Queries about problems with these variables should be directed to the User Support Group

([email protected])


Use of illegal drugs (Based on work undertaken by:

• Richard Rowe & Barbara Maughan, Institute of Psychiatry • Samantha Parsons, Centre for Longitudinal Studies, Institute of Education, University

of London) 1. Frequencies for reported use of illegal drugs look plausible (see table below). As expected, men and BCS70 cohort members reported higher levels of drug use.

Type of drug Taken drug, not in last year Taken drug in last year Cannabis 28.6 11.7 Ecstasy 6.7 2.4 Amphetamines 12.9 2.5 LSD 8.8 0.5 Popper 11.5 1.1 Magic mushrooms 9.7 0.5 Cocaine 5.7 3.4 Temazepan 3.5 0.9 Ketamine 1.1 0.2 Crack 0.9 0.3 Heroin 1.2 0.3 Methadone 0.7 0.2 “Semeron” 0.2 <0.1 Other drugs 2.7

Source: Richard Rowe & Barbara Maughan, Institute of Psychiatry 2. ‘‘Semeron’’ is a fictitious drug which was included among the illegal drugs listed in the self-completion. This same “drug” has been included in many similar instruments in a variety of surveys. Although relatively few cohort members report using semeron, it is clear that some caution should be used when analyzing the data for those who do so. If semeron-users are excluded from the above table, the prevalence of the use of other genuine illegal drugs will be marginally reduced. 3. Additional information about the data relating to illegal drug use is given below. A number of derived variables are also identified. The latter exclude data for cohort members who reported using semeron. Further details of these variables are to be found in the detailed documentation supplied by Samantha Parsons, which is included as Appendix 5.


variable system-missing user-missing: 8 (^

indicates 98) user-missing: 9 (^

indicates 99) valid n

n % n % n % CANNABIS 283 1.2 13 0.1 6 0.0 22378 ECSTASY 283 1.2 12 0.1 7 0.0 22378 AMPHET 283 1.2 11 0 6 0.0 22380 LSD 283 1.2 11 0 6 0.0 22380 POPPER 283 1.2 11 0 8 0.0 22378 MAGMUSH 283 1.2 11 0 7 0.0 22379 COCAINE 283 1.2 11 0 7 0.0 22379 TEMAZ 283 1.2 11 0 9 0.0 22377 SEMERON 283 1.2 11 0 10 0.0 22376 KETAMINE 283 1.2 11 0 8 0.0 22378 CRACK 283 1.2 11 0 6 0.0 22380 HEROIN 283 1.2 11 0 6 0.0 22380 METHAD 283 1.2 11 0 6 0.0 22380 OTHDRUG 283 1.2 11 0 7 0.0 22379 DRUG-DRUG6: multi-coded string variables for the 614 CMs who answered ‘yes’ to othdrug. 22378 are ‘numeric missing’ leaving just 302 individual answers in drug, 37 in drug2, 13 in drug3, 8 in drug4, 4 in drug5, 2 in drug6. Each answer needs to be converted from string to numeric.


Derived ILLEGAL DRUG variables variable system-missing user-missing: 8 (^



n % N % n % If CMs said they had taken SEMERON they were omitted from derived variables on illegal drug use. 43 (0.2%) said they had taken the drug SEMERON. anydrug1 347 1.5 22333 Takes information from the 12 individual illegal drug questions (information in othdrug not coded yet). Counts the number of illegal drugs CM has taken in the last 12 months. 13.5% of CMs have taken at least 1 illegal drug in last 12 months. drugs1 347 1.5 22333 Collapses anydrug1 variable into 0’not taken any drug’ 1’yes, taken at least 1 illegal drug in last 12 months’. anydrug2 347 1.5 22333 Takes information from the 12 individual illegal drug questions (information in othdrug not coded yet). Counts the number of illegal drugs CM has ever taken. 43.5% of CMs have taken at least 1 illegal drug in their lifetime. drugs2 347 1.5 22333 Collapses anydrug2 variable into 0’never taken an illegal drug’ 1’yes, taken at least 1 illegal drug in lifetime’.

Source: Samantha Parsons, Centre for Longitudinal Studies, Institute of Education, University of London. 4. The variables DRUG to DRUG6 hold the ‘Other drugs’ keyed into the laptops by cohort members as they answered the self-completion. Some entries may appear nonsensical, and many are misspelled. Other entries could, in principle be assigned to one of the above categories, (eg: variants on spelling of cannabis etc). A number of other drugs not listed in the self-completion are also identified (eg: opium and solvents). Other entries may well represent street-names of particular drugs. It is hoped that it may be possible to update the variables relating to illegal drugs in order to incorporate this information in the future.



APPENDICES

Page Appendix 1 Pregnancy History Variables A2 Appendix 2 Proxy Interview Variables A5 Appendix 3 NCDS/BCS70 Advisors contributing to quality assessment of NCDS/

BCS70 data A10

Appendix 4 Earnings and related variables A11 Appendix 5 Basic Skills and other variables A57 Appendix 6 Problems with reporting of GCSEs A96 Appendix 7 Highest Qualification A138 Appendix 8 Attitude Scales A163

NCDS/BCS70 1999-2000 Follow-ups: Guide to the Combined Dataset (June 2001) APPENDICES – A1

APPENDIX 1: Pregnancy History Variables

Pregnancy 1 2 Baby 1 2 3 4 5 1 2 3 4 5

Has CM ever been or got anyone else pregnant EVERPREG (Number of babies carried PREGNUM PREGNUM2 How long had CM & partner not been using birth control PREGJ PREGJ2 Months not been using birth control PREGK PREGK2 Weeks not been using birth control PREGKW PREGKW2 Years not been using birth control PREGL PREGL2 MC:Whether smoked during pregnancy CGPREGA1 CGPREGA4 MC:Whether smoked during pregnancy CGPREGA2 CGPREGA5 MC:Whether smoked during pregnancy CGPREGA3 CGPREGA6 Whether smoked more, less or same as before CPREGB CPREGB2 Outcome of pregnancy for baby(n) PREGA PREGA2 PREGA3 PREGA4 PREGA5 PREGA6 PREGA7 PREGA8 PREGA9 PREGA10 Name of baby(n) PREGB PREGB2 PREGB3 PREGB4 PREGB5 PREGB6 PREGB7 PREGB8 PREGB9 PREGB10 Sex of baby(n) PREGC PREGC2 PREGC3 PREGC4 PREGC5 PREGC6 PREGC7 PREGC8 PREGC9 PREGC10 Birth weight of baby(n) – units PREGD PREGD2 PREGD3 PREGD4 PREGD5 PREGD6 PREGD7 PREGD8 PREGD9 PREGD10 Birth weight of baby(n) pounds POUND POUND2 POUND3 POUND4 POUND5 POUND6 POUND7 POUND8 POUND9 POUND10 Birth weight of baby(n) ounces OUNCE OUNCE2 OUNCE3 OUNCE4 OUNCE5 OUNCE6 OUNCE7 OUNCE8 OUNCE9 OUNCE10 Birth weight of baby(n) kilos KILO KILO2 KILO3 KILO4 KILO5 KILO6 KILO7 KILO8 KILO9 KILO10 Birth weight of baby(n) grammes GRAMM GRAMM2 GRAMM3 GRAMM4 GRAMM5 GRAMM6 GRAMM7 GRAMM8 GRAMM9 GRAMM10 baby(n) day of live or still birth PREGED PREGED2 PREGED3 PREGED4 PREGED5 PREGED6 PREGED7 PREGED8 PREGED9 PREGED10 baby(n) mnth of live/still birth, miscarriage/termination PREGEM PREGEM2 PREGEM3 PREGEM4 PREGEM5 PREGEM6 PREGEM7 PREGEM8 PREGEM9 PREGEM10 baby(n) year of live/still birth, miscarriage/termination PREGEY PREGEY2 PREGEY3 PREGEY4 PREGEY5 PREGEY6 PREGEY7 PREGEY8 PREGEY9 PREGEY10 baby(n) date of birth PREGE PREGE2 PREGE3 PREGE4 PREGE5 PREGE6 PREGE7 PREGE8 PREGE9 PREGE10 Was baby(n) early, late or on time PREGF PREGF2 PREGF3 PREGF4 PREGF5 PREGF6 PREGF7 PREGF8 PREGF9 PREGF10 How many weeks early or late was baby(n) PREGG PREGG2 PREGG3 PREGG4 PREGG5 PREGG6 PREGG7 PREGG8 PREGG9 PREGG10 Was anything wrong with baby(n) at birth PREGH PREGH2 PREGH3 PREGH4 PREGH5 PREGH6 PREGH7 PREGH8 PREGH9 PREGH10 What was the problem with baby(n) PREGI PREGI2 PREGI3 PREGI4 PREGI5 PREGI6 PREGI7 PREGI8 PREGI9 PREGI10 Is CMs current partner baby(n)s other parent WHOPARA WHOPARA2 WHOPARA3 WHOPARA4 WHOPARA5 WHOPARA6 WHOPARA7 WHOPARA8 WHOPARA9 WHOPAR10 Who is baby(n)s other parent WHOPARB WHOPARB2 WHOPARB3 WHOPARB4 WHOPARB5 WHOPARB6 WHOPARB7 WHOPARB8 WHOPARB9 WHOPAR11 Does baby(n) live with CM WHERKID WHERKID2 WHERKID3 WHERKID4 WHERKID5 WHERKID6 WHERKID7 WHERKID8 WHERKID9 WHERKI10 What is baby(n) currently doing WHATKID WHATKID2 WHATKID3 WHATKID4 WHATKID5 WHATKID6 WHATKID7 WHATKID8 WHATKID9 WHATKI10 Does baby(n) ever see other parent ABSPARA ABSPARA2 ABSPARA3 ABSPARA4 ABSPARA5 ABSPARA6 ABSPARA7 ABSPARA8 ABSPARA9 ABSPAR10 How often does baby(n) see other parent ABSPARB ABSPARB2 ABSPARB3 ABSPARB4 ABSPARB5 ABSPARB6 ABSPARB7 ABSPARB8 ABSPARB9 ABSPAR11 Does the other parent of baby(n) pay maintenance ABSPARC ABSPARC2 ABSPARC3 ABSPARC4 ABSPARC5 ABSPARC6 ABSPARC7 ABSPARC8 ABSPARC9 ABSPAR12 Where is baby(n) living now ABSKIDA ABSKIDA2 ABSKIDA3 ABSKIDA4 ABSKIDA5 ABSKIDA6 ABSKIDA7 ABSKIDA8 ABSKIDA9 ABSKID10 Has baby(n) ever lived with CM ABSKIDB ABSKIDB2 ABSKIDB3 ABSKIDB4 ABSKIDB5 ABSKIDB6 ABSKIDB7 ABSKIDB8 ABSKIDB9 ABSKID11 Year baby(n) last lived with CM ABSYR ABSYR2 ABSYR3 ABSYR4 ABSYR5 ABSYR6 ABSYR7 ABSYR8 ABSYR9 ABSYR10 Month baby(n) last lived with CM ABSMON ABSMON2 ABSMON3 ABSMON4 ABSMON5 ABSMON6 ABSMON7 ABSMON8 ABSMON9 ABSMON10 Does CM ever see baby(n) now ABSKIDC ABSKIDC2 ABSKIDC3 ABSKIDC4 ABSKIDC5 ABSKIDC6 ABSKIDC7 ABSKIDC8 ABSKIDC9 ABSKID12 How often does CM see baby(n) ABSKIDD ABSKIDD2 ABSKIDD3 ABSKIDD4 ABSKIDD5 ABSKIDD6 ABSKIDD7 ABSKIDD8 ABSKIDD9 ABSKID13 Does CM pay maintenance for baby(n) ABSKIDE ABSKIDE2 ABSKIDE3 ABSKIDE4 ABSKIDE5 ABSKIDE6 ABSKIDE7 ABSKIDE8 ABSKIDE9 ABSKID14 Another pregnancy before this one? MOREPREG MOREPRE2


Pregnancy History Variables (continued)

Pregnancy 3 4 Baby 1 2 3 4 5 1 2 3 4 5

Has CM ever been or got anyone else pregnant (Number of babies carried PREGNUM3 PREGNUM4 How long had CM & partner not been using birth control PREGJ3 PREGJ4 Months not been using birth control PREGK3 PREGK4 Weeks not been using birth control PREGKW3 PREGKW4 Years not been using birth control PREGL3 PREGL4 MC:Whether smoked during pregnancy CGPREGA7 CGPREG10 MC:Whether smoked during pregnancy CGPREGA8 CGPREG11 MC:Whether smoked during pregnancy CGPREGA9 CGPREG12 Whether smoked more, less or same as before CPREGB3 CPREGB4 Outcome of pregnancy for baby(n) PREGA11 PREGA12 PREGA13 PREGA14 PREGA15 PREGA16 PREGA17 PREGA18 PREGA19 PREGA20 Name of baby(n) PREGB11 PREGB12 PREGB13 PREGB14 PREGB15 PREGB16 PREGB17 PREGB18 PREGB19 PREGB20 Sex of baby(n) PREGC11 PREGC12 PREGC13 PREGC14 PREGC15 PREGC16 PREGC17 PREGC18 PREGC19 PREGC20 Birth weight of baby(n) – units PREGD11 PREGD12 PREGD13 PREGD14 PREGD15 PREGD16 PREGD17 PREGD18 PREGD19 PREGD20 Birth weight of baby(n) pounds POUND11 POUND12 POUND13 POUND14 POUND15 POUND16 POUND17 POUND18 POUND19 POUND20 Birth weight of baby(n) ounces OUNCE11 OUNCE12 OUNCE13 OUNCE14 OUNCE15 OUNCE16 OUNCE17 OUNCE18 OUNCE19 OUNCE20 Birth weight of baby(n) kilos KILO11 KILO12 KILO13 KILO14 KILO15 KILO16 KILO17 KILO18 KILO19 KILO20 Birth weight of baby(n) grammes GRAMM11 GRAMM12 GRAMM13 GRAMM14 GRAMM15 GRAMM16 GRAMM17 GRAMM18 GRAMM19 GRAMM20 baby(n) day of live or still birth PREGED11 PREGED12 PREGED13 PREGED14 PREGED15 PREGED16 PREGED17 PREGED18 PREGED19 PREGED20 baby(n) mnth of live/still birth, miscarriage/termination PREGEM11 PREGEM12 PREGEM13 PREGEM14 PREGEM15 PREGEM16 PREGEM17 PREGEM18 PREGEM19 PREGEM20 baby(n) year of live/still birth, miscarriage/termination PREGEY11 PREGEY12 PREGEY13 PREGEY14 PREGEY15 PREGEY16 PREGEY17 PREGEY18 PREGEY19 PREGEY20 baby(n) date of birth PREGE11 PREGE12 PREGE13 PREGE14 PREGE15 PREGE16 PREGE17 PREGE18 PREGE19 PREGE20 Was baby(n) early, late or on time PREGF11 PREGF12 PREGF13 PREGF14 PREGF15 PREGF16 PREGF17 PREGF18 PREGF19 PREGF20 How many weeks early or late was baby(n) PREGG11 PREGG12 PREGG13 PREGG14 PREGG15 PREGG16 PREGG17 PREGG18 PREGG19 PREGG20 Was anything wrong with baby(n) at birth PREGH11 PREGH12 PREGH13 PREGH14 PREGH15 PREGH16 PREGH17 PREGH18 PREGH19 PREGH20 What was the problem with baby(n) PREGI11 PREGI12 PREGI13 PREGI14 PREGI15 PREGI16 PREGI17 PREGI18 PREGI19 PREGI20 Is CMs current partner baby(n)s other parent WHOPAR12 WHOPAR14 WHOPAR16 WHOPAR18 WHOPAR20 WHOPAR22 WHOPAR24 WHOPAR26 WHOPAR28 WHOPAR30 Who is baby(n)s other parent WHOPAR13 WHOPAR15 WHOPAR17 WHOPAR19 WHOPAR21 WHOPAR23 WHOPAR25 WHOPAR27 WHOPAR29 WHOPAR31 Does baby(n) live with CM WHERKI11 WHERKI12 WHERKI13 WHERKI14 WHERKI15 WHERKI16 WHERKI17 WHERKI18 WHERKI19 WHERKI20 What is baby(n) currently doing WHATKI11 WHATKI12 WHATKI13 WHATKI14 WHATKI15 WHATKI16 WHATKI17 WHATKI18 WHATKI19 WHATKI20 Does baby(n) ever see other parent ABSPAR13 ABSPAR16 ABSPAR19 ABSPAR22 ABSPAR25 ABSPAR28 ABSPAR31 ABSPAR34 ABSPAR37 ABSPAR40 How often does baby(n) see other parent ABSPAR14 ABSPAR17 ABSPAR20 ABSPAR23 ABSPAR26 ABSPAR29 ABSPAR32 ABSPAR35 ABSPAR38 ABSPAR41 Does the other parent of baby(n) pay maintenance ABSPAR15 ABSPAR18 ABSPAR21 ABSPAR24 ABSPAR27 ABSPAR30 ABSPAR33 ABSPAR36 ABSPAR39 ABSPAR42 Where is baby(n) living now ABSKID15 ABSKID20 ABSKID25 ABSKID30 ABSKID35 ABSKID40 ABSKID45 ABSKID50 ABSKID55 ABSKID60 Has baby(n) ever lived with CM ABSKID16 ABSKID21 ABSKID26 ABSKID31 ABSKID36 ABSKID41 ABSKID46 ABSKID51 ABSKID56 ABSKID61 Year baby(n) last lived with CM ABSYR11 ABSYR12 ABSYR13 ABSYR14 ABSYR15 ABSYR16 ABSYR17 ABSYR18 ABSYR19 ABSYR20 Month baby(n) last lived with CM ABSMON11 ABSMON12 ABSMON13 ABSMON14 ABSMON15 ABSMON16 ABSMON17 ABSMON18 ABSMON19 ABSMON20 Does CM ever see baby(n) now ABSKID17 ABSKID22 ABSKID27 ABSKID32 ABSKID37 ABSKID42 ABSKID47 ABSKID52 ABSKID57 ABSKID62 How often does CM see baby(n) ABSKID18 ABSKID23 ABSKID28 ABSKID33 ABSKID38 ABSKID43 ABSKID48 ABSKID53 ABSKID58 ABSKID63 Does CM pay maintenance for baby(n) ABSKID19 ABSKID24 ABSKID29 ABSKID34 ABSKID39 ABSKID44 ABSKID49 ABSKID54 ABSKID59 ABSKID64 Another pregnancy before this one? MOREPRE3 MOREPRE4



Pregnancy 5 6 Baby 1 2 3 4 5 1 2 3 4 5

Has CM ever been or got anyone else pregnant (Number of babies carried PREGNUM5 PREGNUM6 How long had CM & partner not been using birth control PREGJ5 PREGJ6 Months not been using birth control PREGK5 PREGK6 Weeks not been using birth control PREGKW5 PREGKW6 Years not been using birth control PREGL5 PREGL6 MC:Whether smoked during pregnancy CGPREG13 CGPREG16 MC:Whether smoked during pregnancy CGPREG14 CGPREG17 MC:Whether smoked during pregnancy CGPREG15 CGPREG18 Whether smoked more, less or same as before CPREGB5 CPREGB6 Outcome of pregnancy for baby(n) PREGA21 PREGA22 PREGA23 PREGA24 PREGA25 PREGA26 PREGA27 PREGA28 PREGA29 PREGA30 Name of baby(n) PREGB21 PREGB22 PREGB23 PREGB24 PREGB25 PREGB26 PREGB27 PREGB28 PREGB29 PREGB30 Sex of baby(n) PREGC21 PREGC22 PREGC23 PREGC24 PREGC25 PREGC26 PREGC27 PREGC28 PREGC29 PREGC30 Birth weight of baby(n) – units PREGD21 PREGD22 PREGD23 PREGD24 PREGD25 PREGD26 PREGD27 PREGD28 PREGD29 PREGD30 Birth weight of baby(n) pounds POUND21 POUND22 POUND23 POUND24 POUND25 POUND26 POUND27 POUND28 POUND29 POUND30 Birth weight of baby(n) ounces OUNCE21 OUNCE22 OUNCE23 OUNCE24 OUNCE25 OUNCE26 OUNCE27 OUNCE28 OUNCE29 OUNCE30 Birth weight of baby(n) kilos KILO21 KILO22 KILO23 KILO24 KILO25 KILO26 KILO27 KILO28 KILO29 KILO30 Birth weight of baby(n) grammes GRAMM21 GRAMM22 GRAMM23 GRAMM24 GRAMM25 GRAMM26 GRAMM27 GRAMM28 GRAMM29 GRAMM30 baby(n) day of live or still birth PREGED21 PREGED22 PREGED23 PREGED24 PREGED25 PREGED26 PREGED27 PREGED28 PREGED29 PREGED30 baby(n) mnth of live/still birth, miscarriage/termination PREGEM21 PREGEM22 PREGEM23 PREGEM24 PREGEM25 PREGEM26 PREGEM27 PREGEM28 PREGEM29 PREGEM30 baby(n) year of live/still birth, miscarriage/termination PREGEY21 PREGEY22 PREGEY23 PREGEY24 PREGEY25 PREGEY26 PREGEY27 PREGEY28 PREGEY29 PREGEY30 baby(n) date of birth PREGE21 PREGE22 PREGE23 PREGE24 PREGE25 PREGE26 PREGE27 PREGE28 PREGE29 PREGE30 Was baby(n) early, late or on time PREGF21 PREGF22 PREGF23 PREGF24 PREGF25 PREGF26 PREGF27 PREGF28 PREGF29 PREGF30 How many weeks early or late was baby(n) PREGG21 PREGG22 PREGG23 PREGG24 PREGG25 PREGG26 PREGG27 PREGG28 PREGG29 PREGG30 Was anything wrong with baby(n) at birth PREGH21 PREGH22 PREGH23 PREGH24 PREGH25 PREGH26 PREGH27 PREGH28 PREGH29 PREGH30 What was the problem with baby(n) PREGI21 PREGI22 PREGI23 PREGI24 PREGI25 PREGI26 PREGI27 PREGI28 PREGI29 PREGI30 Is CMs current partner baby(n)s other parent WHOPAR32 WHOPAR34 WHOPAR36 WHOPAR38 WHOPAR40 WHOPAR42 WHOPAR44 WHOPAR46 WHOPAR48 WHOPAR50 Who is baby(n)s other parent WHOPAR33 WHOPAR35 WHOPAR37 WHOPAR39 WHOPAR41 WHOPAR43 WHOPAR45 WHOPAR47 WHOPAR49 WHOPAR51 Does baby(n) live with CM WHERKI21 WHERKI22 WHERKI23 WHERKI24 WHERKI25 WHERKI26 WHERKI27 WHERKI28 WHERKI29 WHERKI30 What is baby(n) currently doing WHATKI21 WHATKI22 WHATKI23 WHATKI24 WHATKI25 WHATKI26 WHATKI27 WHATKI28 WHATKI29 WHATKI30 Does baby(n) ever see other parent ABSPAR43 ABSPAR46 ABSPAR49 ABSPAR52 ABSPAR55 ABSPAR58 ABSPAR61 ABSPAR64 ABSPAR67 ABSPAR70 How often does baby(n) see other parent ABSPAR44 ABSPAR47 ABSPAR50 ABSPAR53 ABSPAR56 ABSPAR59 ABSPAR62 ABSPAR65 ABSPAR68 ABSPAR71 Does the other parent of baby(n) pay maintenance ABSPAR45 ABSPAR48 ABSPAR51 ABSPAR54 ABSPAR57 ABSPAR60 ABSPAR63 ABSPAR66 ABSPAR69 ABSPAR72 Where is baby(n) living now ABSKID65 ABSKID70 ABSKID75 ABSKID80 ABSKID85 ABSKID90 ABSKID95 ABSKI100 ABSKI105 ABSKI110 Has baby(n) ever lived with CM ABSKID66 ABSKID71 ABSKID76 ABSKID81 ABSKID86 ABSKID91 ABSKID96 ABSKI101 ABSKI106 ABSKI111 Year baby(n) last lived with CM ABSYR21 ABSYR22 ABSYR23 ABSYR24 ABSYR25 ABSYR26 ABSYR27 ABSYR28 ABSYR29 ABSYR30 Month baby(n) last lived with CM ABSMON21 ABSMON22 ABSMON23 ABSMON24 ABSMON25 ABSMON26 ABSMON27 ABSMON28 ABSMON29 ABSMON30 Does CM ever see baby(n) now ABSKID67 ABSKID72 ABSKID77 ABSKID82 ABSKID87 ABSKID92 ABSKID97 ABSKI102 ABSKI107 ABSKI112 How often does CM see baby(n) ABSKID68 ABSKID73 ABSKID78 ABSKID83 ABSKID88 ABSKID93 ABSKID98 ABSKI103 ABSKI108 ABSKI113 Does CM pay maintenance for baby(n) ABSKID69 ABSKID74 ABSKID79 ABSKID84 ABSKID89 ABSKID94 ABSKID99 ABSKI104 ABSKI109 ABSKI114 Another pregnancy before this one? MOREPRE5 MOREPRE6



Pregnancy 7 8 Baby 1 2 3 4 5 1 2 3 4 5

Has CM ever been or got anyone else pregnant (Number of babies carried PREGNUM7 PREGNUM8 How long had CM & partner not been using birth control PREGJ7 PREGJ8 Months not been using birth control PREGK7 PREGK8 Weeks not been using birth control PREGKW7 PREGKW8 Years not been using birth control PREGL7 PREGL8 MC:Whether smoked during pregnancy CGPREG19 CGPREG22 MC:Whether smoked during pregnancy CGPREG20 CGPREG23 MC:Whether smoked during pregnancy CGPREG21 CGPREG24 Whether smoked more, less or same as before CPREGB7 CPREGB8 Outcome of pregnancy for baby(n) PREGA31 PREGA32 PREGA33 PREGA34 PREGA35 PREGA36 PREGA37 PREGA38 PREGA39 PREGA40 Name of baby(n) PREGB31 PREGB32 PREGB33 PREGB34 PREGB35 PREGB36 PREGB37 PREGB38 PREGB39 PREGB40 Sex of baby(n) PREGC31 PREGC32 PREGC33 PREGC34 PREGC35 PREGC36 PREGC37 PREGC38 PREGC39 PREGC40 Birth weight of baby(n) - units PREGD31 PREGD32 PREGD33 PREGD34 PREGD35 PREGD36 PREGD37 PREGD38 PREGD39 PREGD40 Birth weight of baby(n) pounds POUND31 POUND32 POUND33 POUND34 POUND35 POUND36 POUND37 POUND38 POUND39 POUND40 Birth weight of baby(n) ounces OUNCE31 OUNCE32 OUNCE33 OUNCE34 OUNCE35 OUNCE36 OUNCE37 OUNCE38 OUNCE39 OUNCE40 Birth weight of baby(n) kilos KILO31 KILO32 KILO33 KILO34 KILO35 KILO36 KILO37 KILO38 KILO39 KILO40 Birth weight of baby(n) grammes GRAMM31 GRAMM32 GRAMM33 GRAMM34 GRAMM35 GRAMM36 GRAMM37 GRAMM38 GRAMM39 GRAMM40 baby(n) day of live or still birth PREGED31 PREGED32 PREGED33 PREGED34 PREGED35 PREGED36 PREGED37 PREGED38 PREGED39 PREGED40 baby(n) mnth of live/still birth, miscarriage/termination PREGEM31 PREGEM32 PREGEM33 PREGEM34 PREGEM35 PREGEM36 PREGEM37 PREGEM38 PREGEM39 PREGEM40 baby(n) year of live/still birth, miscarriage/termination PREGEY31 PREGEY32 PREGEY33 PREGEY34 PREGEY35 PREGEY36 PREGEY37 PREGEY38 PREGEY39 PREGEY40 baby(n) date of birth PREGE31 PREGE32 PREGE33 PREGE34 PREGE35 PREGE36 PREGE37 PREGE38 PREGE39 PREGE40 Was baby(n) early, late or on time PREGF31 PREGF32 PREGF33 PREGF34 PREGF35 PREGF36 PREGF37 PREGF38 PREGF39 PREGF40 How many weeks early or late was baby(n) PREGG31 PREGG32 PREGG33 PREGG34 PREGG35 PREGG36 PREGG37 PREGG38 PREGG39 PREGG40 Was anything wrong with baby(n) at birth PREGH31 PREGH32 PREGH33 PREGH34 PREGH35 PREGH36 PREGH37 PREGH38 PREGH39 PREGH40 What was the problem with baby(n) PREGI31 PREGI32 PREGI33 PREGI34 PREGI35 PREGI36 PREGI37 PREGI38 PREGI39 PREGI40 Is CMs current partner baby(n)s other parent WHOPAR52 WHOPAR54 WHOPAR56 WHOPAR58 WHOPAR60 WHOPAR62 WHOPAR64 WHOPAR66 WHOPAR68 WHOPAR70 Who is baby(n)s other parent WHOPAR53 WHOPAR55 WHOPAR57 WHOPAR59 WHOPAR61 WHOPAR63 WHOPAR65 WHOPAR67 WHOPAR69 WHOPAR71 Does baby(n) live with CM WHERKI31 WHERKI32 WHERKI33 WHERKI34 WHERKI35 WHERKI36 WHERKI37 WHERKI38 WHERKI39 WHERKI40 What is baby(n) currently doing WHATKI31 WHATKI32 WHATKI33 WHATKI34 WHATKI35 WHATKI36 WHATKI37 WHATKI38 WHATKI39 WHATKI40 Does baby(n) ever see other parent ABSPAR73 ABSPAR76 ABSPAR79 ABSPAR82 ABSPAR85 ABSPAR88 ABSPAR91 ABSPAR94 ABSPAR97 ABSPA100 How often does baby(n) see other parent ABSPAR74 ABSPAR77 ABSPAR80 ABSPAR83 ABSPAR86 ABSPAR89 ABSPAR92 ABSPAR95 ABSPAR98 ABSPA101 Does the other parent of baby(n) pay maintenance ABSPAR75 ABSPAR78 ABSPAR81 ABSPAR84 ABSPAR87 ABSPAR90 ABSPAR93 ABSPAR96 ABSPAR99 ABSPA102 Where is baby(n) living now ABSKI115 ABSKI120 ABSKI125 ABSKI130 ABSKI135 ABSKI140 ABSKI145 ABSKI150 ABSKI155 ABSKI160 Has baby(n) ever lived with CM ABSKI116 ABSKI121 ABSKI126 ABSKI131 ABSKI136 ABSKI141 ABSKI146 ABSKI151 ABSKI156 ABSKI161 Year baby(n) last lived with CM ABSYR31 ABSYR32 ABSYR33 ABSYR34 ABSYR35 ABSYR36 ABSYR37 ABSYR38 ABSYR39 ABSYR40 Month baby(n) last lived with CM ABSMON31 ABSMON32 ABSMON33 ABSMON34 ABSMON35 ABSMON36 ABSMON37 ABSMON38 ABSMON39 ABSMON40 Does CM ever see baby(n) now ABSKI117 ABSKI122 ABSKI127 ABSKI132 ABSKI137 ABSKI142 ABSKI147 ABSKI152 ABSKI157 ABSKI162 How often does CM see baby(n) ABSKI118 ABSKI123 ABSKI128 ABSKI133 ABSKI138 ABSKI143 ABSKI148 ABSKI153 ABSKI158 ABSKI163 Does CM pay maintenance for baby(n) ABSKI119 ABSKI124 ABSKI129 ABSKI134 ABSKI139 ABSKI144 ABSKI149 ABSKI154 ABSKI159 ABSKI164 Another pregnancy before this one? MOREPRE7 MOREPRE8


APPENDIX 2: Proxy Interview Variables Where the cohort member was unable to understand or respond to questions put by the interviewer or to the self-completion, short proxy interviews were undertaken with a family member or carer. The variables which hold the data for the proxy interview are identifiable through their labels – all begin with the endorsement “(Proxy)”. These variables and associated labels are listed below.

Variable Label Sequential

position PROXTHK (Proxy) Thank you for taking part in this interview 31 PROXYWHO (Proxy) What is your relationship to CM 50 PROXYAC1 (Proxy) MC: CM been married, had children, a job or Ed? 51 PROXYAC2 (Proxy) MC: CM been married, had children, a job or Ed? 52 PROXYAC3 (Proxy) MC: CM been married, had children, a job or Ed? 53 PROXYAC4 (Proxy) MC CM been married, had children, a job or Ed? 54 LSIIMWK (Proxy) (LSI1) Illness/disability that limits work CM can do 186 LSIAGE (Proxy) (LSI1) Age when condition started 187 LSILIM (Proxy) (LSI1) Does illness limit activities? 188 MORE11 (Proxy) (LSI1) Other illness, disability or infirmity 189 LSIIMWK2 (Proxy) (LSI2) Illness/disability that limits work CM can do 190 LSIAGE2 (Proxy) (LSI2) Age when condition started 191 LSILIM2 (Proxy) (LSI2) Does CM condition limit activities? 192 MORE12 (Proxy) (LSI2) Does CM have any other condition? 193 LSIIMWK3 (Proxy) (LSI3) Illness/disability that limits work CM can do 194 LSIAGE3 (Proxy) (LSI3) Age when condition started 195 LSILIM3 (Proxy) (LSI3) Does CM condition limit activities? 196 MORE13 (Proxy) (LSI3) Does CM have any other condition? 197 LSIIMWK4 (Proxy) (LSI4) Illness/disability that limits work CM can do 198 LSIAGE4 (Proxy) (LSI4) Age when condition started 199 LSILIM4 (Proxy) (LSI4) Does CM condition limit activities? 200 MORE14 (Proxy) (LSI4) Does CM have any other condition? 201 LSIIMWK5 (Proxy) (LSI5) Illness/disability that limits work CM can do 202 LSIAGE5 (Proxy) (LSI5) Age when condition started 203 LSILIM5 (Proxy) (LSI5) Does CM condition limit activities? 204 MORE15 (Proxy) (LSI5) Does CM have any other condition? 205 LSIIMWK6 (Proxy) (LSI6) Illness/disability that limits work CM can do 206 LSIAGE6 (Proxy) (LSI6) Age when condition started 207 LSILIM6 (Proxy) (LSI6) Does CM condition limit activities? 208 MORE16 (Proxy) (LSI6) Does CM have any other condition? 209 LSIIMWK7 (Proxy) (LSI7) Illness/disability that limits work CM can do 210 LSIAGE7 (Proxy) (LSI7) Age when condition started 211 LSILIM7 (Proxy) (LSI7) Does CM condition limit activities? 212 MORE17 (Proxy) (LSI7) Does CM have any other condition? 213 LSIIMWK8 (Proxy) (LSI8) Illness/disability that limits work CM can do 214 LSIAGE8 (Proxy) (LSI8) Age when condition started 215 LSILIM8 (Proxy) (LSI8) Does CM condition limit activities? 216 MORE18 (Proxy) (LSI8) Does CM have any other condition? 217 LSIIMWK9 (Proxy) (LSI9) Illness/disability that limits work CM can do 218 LSIAGE9 (Proxy) (LSI9) Age when condition started 219 LSILIM9 (Proxy) (LSI9) Does CMs condition limit activities? 220 MORE19 (Proxy) (LSI10) Does CM have any other condition? 221



position LSIIMW10 (Proxy) (LSI0) Illness/disability that limits work CM can do 222 LSIAGE10 (Proxy) (LSI10) Age when condition started 223 LSILIM10 (Proxy) (LSI10) Does CM condition limit activities? 224 MORE20 (Proxy) Does CM have any other condition? 225 HOAGE (Proxy) (HC1)Age when CM first had condition? 226 HOSUP1 (Proxy)(HC1) MC Who monitors this condition 227 HOSUP2 (Proxy) (HC1) MC Who monitors this condition 228 HOSUP3 (Proxy) (HC1) MC Who monitors this condition 229 HOMORE (Proxy) (HC1) Any other health condition 230 HOAGE2 (Proxy) (HC2)Age when CM first had condition? 231 HOSUP4 (Proxy) (HC2) MC Who monitors this condition 232 HOSUP5 (Proxy) (HC2) MC Who monitors this condition 233 HOSUP6 (Proxy) (HC2) MC Who monitors this condition 234 HOMORE2 (Proxy) (HC2) Any other health condition 235 HOAGE3 (Proxy) (HC3)Age when CM first had condition? 236 HOSUP7 (Proxy) (HC3) MC Who monitors this condition 237 HOSUP8 (Proxy) (HC3) MC Who monitors this condition 238 HOSUP9 (Proxy) (HC3) MC Who monitors this condition 239 HOMORE3 (Proxy) (HC3) Any other health condition 240 HOAGE4 (Proxy) (HC4)Age when CM first had condition? 241 HOSUP10 (Proxy) (HC4) MC Who monitors this condition 242 HOSUP11 (Proxy) (HC4) MC Who monitors this condition 243 HOSUP12 (Proxy) (HC4) MC Who monitors this condition 244 HOMORE4 (Proxy) (HC4) Any other health condition 245 HOAGE5 (Proxy) (HC5)Age when CM first had condition? 246 HOSUP13 (Proxy) (HC5) MC Who monitors this condition 247 HOSUP14 (Proxy) (HC5) MC Who monitors this condition 248 HOSUP15 (Proxy) (HC5) MC Who monitors this condition 249 HOMORE5 (Proxy) (HC5) Any other health condition 250 HOAGE6 (Proxy) (HC6)Age when CM first had condition? 251 HOSUP16 (Proxy) (HC6) MC Who monitors this condition 252 HOSUP17 (Proxy) (HC6) MC Who monitors this condition 253 HOSUP18 (Proxy) (HC6) MC Who monitors this condition 254 HOMORE6 (Proxy) (HC6) Any other health condition 255 HOAGE7 (Proxy) (HC7)Age when CM first had condition? 256 HOSUP19 (Proxy) (HC7) MC Who monitors this condition 257 HOSUP20 (Proxy) (HC7) MC Who monitors this condition 258 HOSUP21 (Proxy) (HC7) MC Who monitors this condition 259 HOMORE7 (Proxy) (HC7) Any other health condition 260 HOAGE8 (Proxy) (HC8)Age when CM first had condition? 261 HOSUP22 (Proxy) (HC8) MC Who monitors this condition 262 HOSUP23 (Proxy) (HC8) MC Who monitors this condition 263 HOSUP24 (Proxy) (HC8) MC Who monitors this condition 264 HOMORE8 (Proxy) (HC8) Any other health condition 265 HOAGE9 (Proxy) (HC9)Age when CM first had condition? 266 HOSUP25 (Proxy) (HC9) MC Who monitors this condition 267 HOSUP26 (Proxy) (HC9) MC Who monitors this condition 268 HOSUP27 (Proxy) (HC9) MC Who monitors this condition 269 HOMORE9 (Proxy) (HC9) Any other health condition 270



position HOAGE10 (Proxy) (HC10)Age when CM first had condition? 271 HOSUP28 (Proxy) (HC10) MC Who monitors this condition 272 HOSUP29 (Proxy) (HC10) MC Who monitors this condition 273 HOSUP30 (Proxy) (HC10) MC Who monitors this condition 274 HOMORE10 (Proxy) (HC10) Any other health condition 275 YEARIN (Proxy) Year CM moved into current address 276 YEARM (Proxy) Month CM moved into current address 277 TENURE (Proxy) 'Does CM own or rent home' 278 MARSTAT (Proxy) What is CMs legal marital status 279 PREGPROX (Proxy) How many children does CM have 280 PROXYTYP (Proxy)Whether CM was employed or self employed 281 PROXYECO (Proxy) Was CM working FT or PT 282 CJSUP (Proxy)Did CM have managerial duties 283 CJEMPS (Proxy) How many employees were there 284 AGELFTED (Proxy) CMs age when left FT ed 285 ACTAGEL (Proxy) Whether still in FT education 286 FURTHED (Proxy) Whether CM started FT education later 287 LFTMORED (Proxy)Age CM left last period of education 288 EDQUALS (Proxy) Any academic quals obtained since Ref date 289 EDQTYP01 (Proxy)MC: Quals obtained since Ref date 290 EDQTYP03 (Proxy)MC: Quals obtained since Ref date 292 EDQTYP04 (Proxy)MC: Quals obtained since Ref date 293 EDQTYP05 (Proxy)MC: Quals obtained since Ref date 294 EDQTYP06 (Proxy)MC: Quals obtained since Ref date 295 EDQTYP07 (Proxy)MC: Quals obtained since Ref date 296 EDQTYP08 (Proxy)MC: Quals obtained since Ref date 297 EDQTYP09 (Proxy)MC: Quals obtained since Ref date 298 EDQTYP10 (Proxy)MC: Quals obtained since Ref date 299 EDQTYP11 (Proxy)MC: Quals obtained since Ref date 300 EDQTYP12 (Proxy)MC: Quals obtained since Ref date 301 EDQTYP13 (Proxy)MC: Quals obtained since Ref date 302 VOCQUALS (Proxy) Any vocational quals obtained since Refdate 303 VOCTYP01 (Proxy)MC: Voc. quals obtained since Ref date 304 VOCTYP02 (Proxy)MC: Voc. quals obtained since Ref date 305 VOCTYP03 (Proxy)MC: Voc. quals obtained since Ref date 306 VOCTYP04 (Proxy)MC: Voc. quals obtained since Ref date 307 VOCTYP05 (Proxy)MC: Voc. quals obtained since Ref date 308 VOCTYP06 (Proxy)MC: Voc. quals obtained since Ref date 309 VOCTYP07 (Proxy)MC: Voc. quals obtained since Ref date 310 VOCTYP08 (Proxy)MC: Voc. quals obtained since Ref date 311 VOCTYP09 (Proxy)MC: Voc. quals obtained since Ref date 312 VOCTYP10 (Proxy)MC: Voc. quals obtained since Ref date 313 VOCTYP11 (Proxy)MC: Voc. quals obtained since Ref date 314 LSIANY (Proxy) Does CM have long-standing illness/disability? 315 LSIREG (Proxy) Is CM a registered disabled person 316 HEIGHT (Proxy) CM height without shoes - units 318 HTMETRES (Proxy) CM height without shoes - metres 319 HTCMS (Proxy) CM height without shoes - centimetres 320 HTFEET (Proxy) CM height without shoes - feet 321



position HTINCHES (Proxy) CM height without shoes - inches 322 WEIGHT (Proxy) CM current weight without clothes - units 323 WTKILOS (Proxy) CM current weight - kilos 324 WTSTONES (Proxy) CM current weight - stones 325 WTPOUNDS (Proxy) CM current weight - pounds 326 IFPROXY Proxy interview indicator 6661 NMPRX (Proxy) CM name 6662 NMPRX2 (Proxy) Person2 name 6664 NMPRX3 (Proxy) Person3 name 6666 NMPRX4 (Proxy) Person4 name 6668 NMPRX5 (Proxy) Person5 name 6670 NMPRX6 (Proxy) Person6 name 6672 NMPRX7 (Proxy) Person7 name 6674 NMPRX8 (Proxy) Person8 name 6676 NMPRX9 (Proxy) Person9 name 6678 NMPRX10 (Proxy) Person10 name 6680 SEXPRX (Proxy) CM gender 6682 SEXPRX2 (Proxy) Person2 gender 6684 SEXPRX3 (Proxy) Person3 gender 6686 SEXPRX4 (Proxy) Person4 gender 6688 SEXPRX5 (Proxy) Person5 gender 6690 SEXPRX6 (Proxy) Person6 gender 6692 SEXPRX7 (Proxy) Person7 gender 6694 SEXPRX8 (Proxy) Person8 gender 6696 SEXPRX9 (Proxy) Person9 gender 6698 SEXPRX10 (Proxy) Person10 gender 6700 AGEPRX (Proxy) CM age last birthday 6702 AGEPRX2 (Proxy) Person2 age last birthday 6704 AGEPRX3 (Proxy) Person3 age last birthday 6706 AGEPRX4 (Proxy) Person4 age last birthday 6708 AGEPRX5 (Proxy) Person5 age last birthday 6710 AGEPRX6 (Proxy) Person6 age last birthday 6712 AGEPRX7 (Proxy) Person7 age last birthday 6714 AGEPRX8 (Proxy) Person8 age last birthday 6716 AGEPRX9 (Proxy) Person9 age last birthday 6718 AGEPRX10 (Proxy) Person10 age last birthday 6720 MSPRX (Proxy) CM Current partnership status 6722 MSPRX2 (Proxy) Person2 Current partnership status 6724 MSPRX3 (Proxy) Person 3 Current partnership status 6726 MSPRX4 (Proxy) Person4 Current partnership status 6728 MSPRX5 (Proxy) Person5 Current partnership status 6730 MSPRX6 (Proxy) Person6 Current partnership status 6732 MSPRX7 (Proxy) Person7 Current partnership status 6734 MSPRX8 (Proxy) Person8 Current partnership status 6736 MSPRX9 (Proxy) Person9 Current partnership status 6738 MSPRX10 (Proxy) Person10 Current partnership status 6740 RELPRX (Proxy) Relationship to CM (always blank for CM) 6742 RELPRX2 (Proxy) Person2 Relationship to CM 6744 RELPRX3 (Proxy) Person3 Relationship to CM 6746



position RELPRX4 (Proxy) Person4 Relationship to CM 6748 RELPRX5 (Proxy) Person5 Relationship to CM 6750 RELPRX6 (Proxy) Person6 Relationship to CM 6752 RELPRX7 (Proxy) Person7 Relationship to CM 6754 RELPRX8 (Proxy) Person8 Relationship to CM 6756 RELPRX9 (Proxy) Person9 Relationship to CM 6758 RELPRX10 (Proxy) Person10 Relationship to CM 6760 CMSEXPX (Proxy) Interviewer check on CM gender 6763 BDAT1PX (Proxy) Check: what date in April is CMs birthday 6765 BDAT2PX (Proxy) Check: what date in Match is CMs birthday 6767 CMNAMPRX (Proxy) CM name from sample file 6769 RENAMPRX (Proxy) Interviewer: record name change on ARF 6771 RESIDPRX (Proxy) Type of accommodation 6774 NORMPRX (Proxy) Does CM usually live at this address 6775


APPENDIX 3: NCDS/BCS70 Advisors contributing to quality assessment of NCDS/BCS70 data A list of the advisors and their associates who contributed to the assessment of the quality of the data, and whose work is drawn on above is given in below. The contribution of individual advisors is acknowledged below Wherever appropriate, the detailed documentation provided by advisors is included in Appendices 4-8.

Dr Ann Berrington University of Southampton Louise Blackwell Centre For Longitudinal Studies, Institute of Education, University of London Lorraine Dearden Institute for Fiscal Studies Alissa Goodman Institute for Fiscal Studies Barbara Jefferis Institute of Child Health Andrew Jenkins Institute of Education, University of London Gerald Makepeace University of Cardiff Scott Montgomery Karolinska Sjukhuset, Stockholm, Sweden Barbara Maughan Institute of Psychiatry Samantha Parsons Centre For Longitudinal Studies, Institute of Education, University of London Dr Gaëlle Pierre Institute for Employment Research John Preston Centre For Longitudinal Studies, Institute of Education, University of London Richard Rowe Institute of Psychiatry


APPENDIX 4: Earnings and related variables Documentation produced by Alissa Goodman and Lorraine Dearden of the Institute for Fiscal Studies relating to earnings and related variables is reproduced below. (NB: References to Appendix 1 and Appendix 2 in the original documentation have been changed below to Annex 1 and Annex 2 to avoid confusion with Appendices 1 and 2 above)

Queries about the contents of this appendix should be directed to the User Support Group

([email protected])

Derived Variables. Section 1. Interview date 1. Corrected interview date Name: Iintdate Description Corrected interview date Purpose To correct interview dates that were originally coded as outside the interviewing period. In the original data there were 23 individuals whose interview dates were not between 10/99 and 9/00. Source variables Intdate Code See intdat.do Section 2. Deflators 2. RPI at interview date Name: rpi_int Description Retail price index (RPI) for month of interview, where Jan 2001 = 1.000. Purpose To allow wages, incomes etc. to be converted into other prices. To uprate wages to Jan 2001 prices, divide by this deflator. Source variables I_intdate, Retail price index (RPI). Frequency distribution See Annex 1, Table 2. Section 3. Employment variables 3. Total usual weekly hours Name: Hours Description Total usual weekly hours, including paid and unpaid overtime, not including meal breaks. Purpose To derive a measure of total hours usually worked, in order to convert gross and net pay information available into an hourly pay measure. Source variables Chours1 Chours2 Chours3 Chours4, Otimeany. This is calculated as: Chours1 + Chours2 + Chours3 + Chours4, where missing values of each of these four variables are set to zero. This ensures that these variables do not include missing values (ie so that no 999s are added in to the total). Notice that individuals have either responses at Chours1 (where Otimeany=2), or responses at Chours2,Chours3 or Chours4 (where Otimeany=1), but never both.


Note this variable is set to missing (.) for all those who a) are not employees b) have missing values in all of the above source variables. Code (Stata do-file format): see checkwage.do Frequency distribution See Annex, Table 3. 4. Imputed/Corrected net last pay Name: Inetpay Description Imputed or corrected net last pay variable. Purpose Clean last net pay variable, corrected for implausible values. Source variables Cnetpay, Cgropay, cnetprd, cgroprd, cnetpred, cgorpred, hours. Source code See checkwage.do Frequency distribution See Annex 1, Table 4. Further information about corrections made: see Annex 2 5. Imputed/Corrected net period Name: Inetprd Description Imputed or corrected net last pay PERIOD variable. Purpose Clean last net pay PERIOD variable, corrected for implausible values and imputed from gross where net unavailable Source variables Cnetpay, Cgropay, cnetprd, cgroprd, cnetpred, cgorpred, hours. Frequency distribution See Annex 1, Table 5. Further information about corrections made: see Annex 2 6. Imputed/Corrected gross pay Name: Igropay Description Imputed or corrected gross last pay variable. Purpose Clean last gross pay variable, corrected for implausible values and imputed from gross where net unavailable Source variables Cnetpay, Cgropay, cnetprd, cgroprd, cnetpred, cgorpred, hours, Source code See checkwage.do Frequency distribution See Annex 1, Table 6. Further information about corrections made: see Annex 2. 7. Imputed/Corrected gross period Name: Igroprd Description Imputed or corrected gross last pay PERIOD variable. Purpose Clean last gross pay PERIOD variable, corrected for implausible values and imputed from gross where net unavailable Source variables Cnetpay, Cgropay, cnetprd, cgroprd, cnetpred, cgorpred, hours, Further information about corrections made: see Annex 2. Frequency distribution


See Annex 1, Table 7. 8. Hourly net last pay Name: hr_net Description: Hourly net last pay. Purpose Clean hourly equivalent last net pay. Where net pay variables are missing but gross pay variables are available, net pay has been imputed from gross using known parameters of the tax system in the relevant year. Source Variables InetPay, InetPrd, Inetpred, Igropay, Igorprd, Igropred, Hours, ms, age2-age10 Code Stata do file format: see checkwage.do, impute.do Frequency distribution See Annex 1, Table 8. 9. Annual net last pay Name: ann_net Description: Clean annual equivalent net last pay. Where net pay variables are missing but gross pay variables are available, net pay has been imputed from gross using known parameters of the tax system in the relevant year. Purpose Clean annual equivalent last net pay, with missing values imputed wherever possible. Source Variables InetPay, InetPrd, Inetpred, Igropay, Igorprd, Igropred, Hours, ms, age2-age10 Code Stata do file format: checkwage.do, impute.do Frequency distribution See Annex 1, Table 9. 10. Hourly gross last pay Name: hr_gro Description: Clean hourly equivalent last gross pay. Where gross pay variables are missing but net pay variables are available, gross pay has been imputed from net using known parameters of the tax system in the relevant year. Purpose Clean hourly equivalent gross last pay, with missing values imputed wherever possible. Source Variables IgroPay, IgroPrd, Igropred, Inetpay, Inetprd,Inetpred Hours, ms, age2-age10 Code Stata do file format: see checkwage.do, impute.do. Frequency distribution See Annex 1, Table 10. 11. Annual gross last pay Name: ann_gro Description: Clean annual equivalent last gross pay. Where gross pay variables are missing but net pay variables are available, gross pay has been imputed from net using known parameters of the tax system in the relevant year. Purpose Clean annual equivalent gross last pay with missing values imputed wherever possible. Source Variables I_groPay, I_groPrd, I_gropred, Inetpay,Inetprd, Ineptpred, Hours, ms, age2-age10 Code Stata do file format: checkwage.do, impute.do Frequency distribution


See Annex 1, Table 11. Section 4. Partner’s earnings variables. 12. Partner’s Imputed/Corrected last net pay Name: Ipnetpay Description: Partner’s last net pay variable, implausible values corrected. Purpose Clean partner’s last net pay. Source Variables Pnetpay, pnetprd, pnetpred Code Stata do file format: partner.do Frequency distribution See Annex 1, Table 12. 13. Partner’s Imputed/Corrected last net pay PERIOD Name: Ipnetprd Description: Partner’s last net pay PERIOD variable, implausible values corrected. Purpose Clean partner’s last net pay PERIOD. Source Variables Pnetpay, pnetprd, pnetpred Code Stata do file format: partner.do Frequency distribution See Annex 1, Table 13. 14. Partner’s Imputed/Corrected last net pay PERIOD Name: Ipnpred Description: Partner’s last net pay PERIOD variable, implausible values corrected. Purpose Clean partner’s last net pay PERIOD. Source Variables Pnetpay, pnetprd, pnetpred Code Stata do file format: partner.do Frequency distribution See Annex 1, Table 14. 15. Partner’s annual last net pay Name: ann_pnet Description: Annual equivalent of partner’s last net pay, implausible values corrected. Purpose Clean annual equivalent partner’s last net pay Source Variables Ipnetpay, Ipnetprd, Ipnetpred Code Stata do file format: partner.do Frequency distribution See Annex 1, Table 15. 16. Partner’s IMPUTED annual last gross pay Name: ann_pgro Description:


Annual equivalent of partner’s last gross pay, imputed from partner’s last net pay using the known parameters of the tax system in the relevant year. Purpose Clean annual equivalent partner’s last gross pay - imputed from net. Source Variables Ipnetpay, Ipnetprd, Ipnetpred, Iintdate, ms Code Stata do file format: partner.do Frequency distribution See Annex 1, Table 16. Section 5. Intermediate variables – used during cleaning. 17. Type of fix to pay variable and period codes Name: fixtype Description Categorical variable containing the type of coding error (if any) which we have identified in the pay information of each employee, and the fix which has been implemented. Purpose To identify and record corrections to last net and gross pay variables. Code See checkwage.do Frequency distribution See Annex 1, Table 17. Further information See Annex 2. 18. Indicator of odd proportion of gross to net pay Name: oddprop Description Indicator of whether ratio of gross to net pay suggests coding error in some of the pay variables which REMAINS UNCORRECTED. Purpose To identify odd wage information Code Oddprop = 1 if (Hr_net/hr_gro) <0.4 or >4 See checkwage.do. See Annex 1, Table 18. Further information, including description and frequency distributions: See Annex 2 19. Indicator of wage variable probably too high or too low Name: oddwage Description Indicator of whether pay variables remain apparently implausibly high or low AFTER CORRECTIONS have been implemented. Purpose To identify odd wage information Code Oddwage =1 id (Hr_gro<2) or (Hr_gro>100) See checkwage.do See Annex 1, Table 19. Further information, including description and frequency distributions: See Annex 2 20. Missing hours variable Name: zhours. Description Indicator of Usual weekly hours variable (hours) missing. Purpose Code See checkwage.do Frequency distribution See Annex 1, Table 20.


21. Missing hourly net pay indicator Name: zhr_net. Description Indicator of if/ why net hourly pay variable is missing. Purpose Code See checkwage.do Frequency distribution See Annex 1, Table 21. 22. Missing hourly gross pay indicator Name: zhr_gro. Description Indicator of if/ why gross hourly pay variable is missing. Purpose Code See checkwage.do Frequency distribution See Annex 1, Table 22. 23. Odd partner’s wage indicator Name: podd. Description Indicator of apparently implausibly high or low partner’s net wage, remaining AFTER CORRECTIONS have been implemented. Purpose Code Podd=1 if (assuming fixed number of hours for part-time and full-time work) implied hourly net pay of partner is <2 or >100. See partner.do Frequency distribution See Annex 1, Table 23.


ANNEX 1 FREQUENCY DISTRIBUTIONS

The frequency distributions are available on request from the User Support Group

([email protected])


ANNEX 2: Explaining IFS suggested fixes to CnetPay, CnetPrd, CgroPay, CgroPrd. Derived variables used for identifying coding errors in the pay information: 1) Fixtype. This is a derived categorical variable taking values 0-20. It contains the type of coding

error (if any) which we have identified in the pay information of each employee, and the suggested fix.

2) Oddprop. This is a (0,1) dummy variable, which is set to 1 if AFTER ALL SUGGESTED CORRECTIONS, the ratio of net to gross hourly pay is still less than 0.4 or greater than 4.

3) Oddwage This is a (0,1) dummy variable, which is set to 1 if AFTER ALL SUGGESTED CORRECTIONS, gross hourly pay is still less than £2 per hour or more than £100 per hour.

How coding errors have been identified In order to identify possible coding errors, we have converted all gross and net pay given to its hourly equivalent (wherever possible – there are a few cases of missing hours variables). We have done this using CnetPay, CnetPrd, Cnetpred, CgroPay, CgroPrd, CgroPred, Chours1-Chours4 (and in a small number of cases where pay is given daily, we also use EconAct). [All our derived variables and code will be deposited in due course.] Some of the coding errors have been spotted because the ratio of net hourly pay to gross hourly pay is clearly wrong. We have considered just those whose ratio of net to gross hourly pay is over 4, or below 0.4. By looking at the data for these observations it has in many cases been fairly obvious what the coding error has been and how it should be corrected. Those with these coding errors (and their suggested solutions) are contained in fixtype 1-12. Those whose ratio of gross to net is wrong, but no obvious fix can be seen are coded as Fixtype=19. Other coding errors have been identified because both gross and net hourly pay are either implausibly high or implausibly low. We have considered those whose implied gross hourly pay is over £100 per hour, or less than £2 per hour. All the coding errors for those whose implied hourly pay is too high or too low are contained in fixtype 13-18. Where it is not clear that a coding error has occurred, or the fix required is not obvious, fixtype is set to 20. Note that these are not necessarily the only coding errors contained in the data. Those considered are those where the ratio of net to gross hourly pay is >4 or <0.4, and those whose implied gross hourly pay is less than £2 or more than £100 per hour. These are arbitrary cut-offs, and a wider net could be considered if needs be. Notice also that a judgement has to be made in the absence of any further information, for those coded in fixtype 8, 14 and fixtype 17 (and possibly others) where a number of different corrections could be made to the data to make it more plausible, and we have just chosen one possible solution. In particular in these cases it is not clear whether the problem is too many (or too few) zeros in the pay amount variables, or whether the period codes are wrong. We have suggested that it is the period codes are wrong, but CLS may take a different view. For those who are coded in fixtype= 19 or fixtype=20, or for whom oddprop=1 or oddwage=1 despite the suggested fixes already put in place, we will have to make some decisions on an individual case-by-case basis whether to leave the information as it is, correct it in some way, or discard it.

The coding frame for Fixtype The table below sets out the coding frame for the fixtype variable: it describes what the miscoding appears to be, and what the suggested correction is. The corrections are simply suggestions; in many cases other possible also corrections exist: see below.


Coding frame for variable fixtype FIXTYPE 0 No fixes made. 1 Gross pay given annually needs to be multiplied by 10. 2 Net pay given monthly needs to be divided by 100. 3 Net pay given monthly needs to be divided by 10. 4 Net pay given monthly or four weekly needs to be multiplied by 100. 5 Net pay given monthly or four weekly needs to be multiplied by 10. 6 Net monthly pay is coded as yearly: CnetPrd needs to be changed to 4. 7 Gross yearly pay is coded as monthly or 4-weekly: CGroPrd needs to be changed to

5. 8* Gross yearly pay coded as weekly: period code needs to be changed to 5. 9 Gross pay given yearly needs to be divided by 10. 10 Gross monthly pay is coded as weekly: CGroPrd needs to be changed to changed to

4. 11 Individual fixes: see code provided below. 12 Gross pay set to missing. 13 Net and Gross pay need to be divided by 100. 14* Net and Gross monthly pay is coded as yearly: CnetPrd and CGroPrd need to be

changed to 4. 15 Net and Gross weekly pay is coded as yearly: CnetPrd and CgroPrd need to be

changed to 4. 16 Total hours variable (derived by IFS) is too low and incompatible with self-reported

economic status. Set hours to missing. 17* Net and Gross yearly pay is coded as monthly: CnetPrd and CgroPrd need to be

changed to 5. 18 Net and Gross pay have both been coded as 1: set to missing. 19 No fix yet: proportion of net to gross pay is odd, but fix needed is not clear. 20 No fix yet: Hourly pay appears very high or low, but fix needed is not clear. *Sometimes there are a number of possible fixes which would make the pay data look more sensible, and it is not entirely clear what the right one is. In particular: Fixtype 8: it has been assumed that the mistake made is that gross annual pay has been coded as weekly. In many cases it is also plausible that it could be gross weekly pay which needs to be divided by 100. Fixtype 14: it has been assumed that the mistake made is that monthly pay has been coded as annual. In fact it could also be the case that it is in fact annual pay which needs to be multiplied by 10 (re-coding to monthly effectively multiplies by 12). Fixtype 17: it has been assumed that the mistake made is that annual pay has been coded as monthly. Equally plausible is that monthly pay has been provided which needs to be divided by 10 (re-coding the period code to annual effectively multiplies it by 12). Details of individual fixes already made. Those coded fixtype=11 require individual fixes, set out below. Individual fixes Serial Fix

Changes to Cgroprd 112006 Cgroprd recoded to 1 101716 Cgroprd recoded to 4 213363 Cgroprd recoded to 4 215256 Cgroprd recoded to 6 Cgropred recoded to 15

Changes to Cnetprd 226017 Cnetprd recoded to 4 229217 Cnetprd recoded to 4 200306 Cnetprd recoded to 1 (although net becomes a bit too high compared to gross) 117261 Cnetprd recoded to 1 (although net becomes a bit too high compared to gross) 219899 Cnetprd recoded to 1 (although net becomes a bit too high compared to gross)


208399 Cnetprd recoded to 1 (although net becomes a bit too high compared to gross) 104694 Cnetprd recoded to 1 (although net becomes a bit too high compared to gross) 105206 Cnetprd recoded to 1 (although net becomes a bit too high compared to gross) 109860 Cnetprd recoded to 2 (gross pay is given fortnightly and this looks more plausible) Changes to cnetpay 119813 Cnetpay should be equal to £100 124299 Cnetpay should be divided by 10 213092 Cnetpay should be multiplied by 10 229425 Cnetpay should be multiplied by 10

Changes to cgropay 121730 Cgropay should be multiplied by 1000 210049 Cgropay should be divided by 10 219518 Cgropay should be divided by 10 228991 Cgropay should be divided by 10 {based on fact that real net weekly pay in NCDS 5

was £605.30 ie. In the region of £600, not £6000 as implied here}

Frequency distribution for fixtype type of fix to net | and/ or gross pay | Freq. Percent Cum. -----------------------+----------------------------------- no fix | 15443 95.20 95.20 gro*10 | 134 0.83 96.03 net/100 | 4 0.02 96.05 net/10 | 17 0.10 96.16 net*100 | 3 0.02 96.18 net*10 | 9 0.06 96.23 net year to month | 167 1.03 97.26 gro month to year | 97 0.60 97.86 gro week to year | 10 0.06 97.92 gro/10 | 15 0.09 98.01 gro week to month | 15 0.09 98.11 individual fix | 22 0.14 98.24 gro to missing | 3 0.02 98.26 both/100 | 14 0.09 98.35 both year to month | 135 0.83 99.18 both year to week | 9 0.06 99.24 hours to missing | 14 0.09 99.32 both month to year | 11 0.07 99.39 no fix- prop still odd | 11 0.07 99.46 no fix- wages still od | 88 0.54 100.00 -----------------------+----------------------------------- Total | 16221 100.00

Further serial numbers under question: oddprop and oddwage After these “corrections” set out above have been carried out, there are still a number of individuals whose wages or ratio of net to gross pay are still odd. We will have to consider these on a case-by-case basis and decide whether to leave the information as it is, correct it in some way, or discard the information all together. The frequency distributions for oddprop and oddwage are as follows. ODDPROP (ratio net: gross hourly pay <0.4 or >4) odd ratio | of net to | gross | hourly pay | Freq. Percent Cum. ------------+----------------------------------- 0 | 16464 99.91 99.91 1 | 15 0.09 100.00 ------------+----------------------------------- Total | 16479 100.00 ODDWAGE (gross hourly wage>£100 or < £2 )


Strangely | high or low | gross hrly | wage | Freq. Percent Cum. ------------+----------------------------------- 0 | 16379 99.39 99.39 1 | 100 0.61 100.00 ------------+----------------------------------- Total | 16479 100.00

Source code For reference, the exact stata code which has been used to derive the fixtype variable is also attached as a stata do-file, checkwage.do.


STATA CODE: • CHECKWAGE.DO • IMPUTE.DO • PARTNER.DO CHECKWAGE.DO /* WAGES DATA */ #delimit; set more 1; /*cd "j:\2000cohorts\derived data"; /* TO MAKE SURE THIS DATASET IS CREATED, RUN GETDATA.DO */ u wagedata,clear; */ /* SET ORIGINAL VARIABLES cnetpay, cnetprd, cgropay, cgroprod IN STONE IN onetpay, ogropay etc. corrections and imputations are then added to cnetpay... */ cap drop onetpay; gen onetpay=cnetpay; lab var onetpay "original net pay variable"; cap drop ogropay; gen ogropay=cgropay; lab var ogropay "original gross pay variable"; cap drop onetprd; gen onetprd=cnetprd; lab var onetprd "original net period variable"; cap drop ogroprd; gen ogroprd=cgroprd; lab var ogroprd "original gross period variable"; /**** EMPLOYMENT STATUS : will be useful later *****************/ cap drop employee; gen employee =1 if econact<=2; replace employee =0 if employee ~=1; lab var employee "in paid employment"; /**************************** INITIAL MISSING GROSS AND NET PAY VARIABLES ******************************/ /* NOTE these will change later, as some nonsense values get set to missing but we need to use them in the interim */ /* missing cnetpay */ cap drop znetpay; gen znetpay =1 if cnetpay==9999998|cnetpay==9999999; replace znetpay =0 if znetpay~=1; replace znetpay =. if cnetpay==.; lab var znetpay "missing cnetpay";


/* missing cgropay */ cap drop zgropay; gen zgropay =1 if cgropay==9999998|cgropay==9999999; replace zgropay =0 if zgropay~=1; replace zgropay =. if cgropay==.; lab var zgropay "missing cgropay"; /***************************** HOURS ******************************************************************/ /* have to set up USUAL WEEKLY HOURS before period codes, to be able to convert pay into hourly */ cap pr drop dohours; pr def dohours; cap drop hours; cap drop zhours; gen zhours = 1 if chours1==. & chours2==. & chours3==. & chours4==. & employee==1; replace zhours = 0 if zhours~=1; replace zhours = . if employee==0; lab var zhours "missing hours"; /* People either get coded at chours1 or at chours2-chours4. This makes it safe to add them all up to get a total, once missing values are taken care of */ /* For chours1, value 99 appears to be missing value. For chours2-chours4, value 999 is missing */ /* Note: a few of these hours variables are set to missing later if they are implausibly low and so mess up the hourly pay variables: see FIX 14 below */ cap drop hours; cap drop hrs1; cap drop hrs2; cap drop hrs3; cap drop hrs4; cap drop paidhrs; gen hrs1= chours1 if chours1<99; gen hrs2= chours2 if chours2<998; gen hrs3= chours3 if chours3<998; gen hrs4= chours4 if chours4<998; replace hrs1 = 0 if hrs1==.; replace hrs2 = 0 if hrs2==.; replace hrs3 = 0 if hrs3==.; replace hrs4 = 0 if hrs4==.; gen hours = hrs1 + hrs2 + hrs3 + hrs4 if zhours==0 & employee==1; gen paidhrs = hrs1 + hrs2 + hrs3 if zhours==0 & employee==1;


/* some have total hours recoreded as zero */ replace zhours = 1 if hours ==0 & employee==1; replace hours = . if zhours ==1 & employee==1; lab var hours "total usual weekly hours"; lab var paidhrs "total paid hours"; gr hours, bin(50) xlab(10,20,30,40,50,60,70,80,90,100,120,140,170) ylab(0,0.05,0.1,0.15,0.2) t1("distribution of hours"); cap log close; log using hours, replace; tab hours; tab zhours; log close; end; dohours; /******************************************* PAY PERIOD CODES *******************************************/ /* the coding frame for CNETPROD and CGROPRD is: 1 one week 2 a fortnight 3 four weeks 4 a calendar month 5 a year 6 other period the coding frame for CNETPRED and CGROPRED is: 1 one week 2 a fortnight/2 weeks 3 four weeks 4 a calendar month 5 a year (12 months/52 weeks) 6 three weeks 7 five weeks 8 six weeks 9 seven weeks 10 eight weeks 11 two calendar months 12 eight times a year 13 nine times a year 14 ten times a year 15 three months/13 weeks 16 six months/26 weeks 17 hourly 18 daily 19 one off lump sum 20 some other period 21 varies 22 refused 23 some other comment 24 irrelevant/unspecific response */


/*********************************************** NET PAY VARIABLES **************************************/ /* First make sure not to start manipulating cnetpay and cgropay missing values 9999997 and 9999998 ! */ replace cnetpay =. if znetpay==1; replace cgropay =. if zgropay==1; /* need to guess how many days per week someone works to convert any daily pay into weekly not too important as it only affects 7 individuals, with a max of 50 hours per week */ /* use paid hours, as we want to sum daily pay over the number of days they get paid for to get total weekly pay */ cap drop days; gen days = paidhrs/8; replace days = 7 if days>=7; lab var days "days per week (estimated)"; /* this program converts net pay into different periods depending on the period code */ cap pr drop donet; pr def donet; cap drop ann_net; cap drop wk_net; cap drop hr_net; cap drop mo_net; cap drop znetprd; gen wk_net = (cnetpay) if cnetprd==1 & znetpay==0; replace wk_net = (cnetpay / 2) if cnetprd==2 & znetpay==0; replace wk_net = (cnetpay / 4) if cnetprd==3 & znetpay==0; replace wk_net = (cnetpay / (13/3)) if cnetprd==4 & znetpay==0; replace wk_net = (cnetpay / 52) if cnetprd==5 & znetpay==0; replace wk_net = 9999997 if cnetprd==. & employee==1 & znetpay==0; replace wk_net = (cnetpay) if cnetprd==6 & cnetpred==1 & znetpay==0; replace wk_net = (cnetpay / 2) if cnetprd==6 & cnetpred==2 & znetpay==0; replace wk_net = (cnetpay / 3) if cnetprd==6 & cnetpred==3 & znetpay==0; replace wk_net = (cnetpay / (13/3)) if cnetprd==6 & cnetpred==4 & znetpay==0;


replace wk_net = (cnetpay / 52) if cnetprd==6 & cnetpred==5 & znetpay==0; replace wk_net = (cnetpay / 3) if cnetprd==6 & cnetpred==6 & znetpay==0; replace wk_net = (cnetpay / 5) if cnetprd==6 & cnetpred==7 & znetpay==0; replace wk_net = (cnetpay / 6) if cnetprd==6 & cnetpred==8 & znetpay==0; replace wk_net = (cnetpay / 7) if cnetprd==6 & cnetpred==9 & znetpay==0; replace wk_net = (cnetpay / 8) if cnetprd==6 & cnetpred==10 & znetpay==0; replace wk_net = (cnetpay / (26/3)) if cnetprd==6 & cnetpred==11 & znetpay==0; replace wk_net = (cnetpay / (52/8)) if cnetprd==6 & cnetpred==12 & znetpay==0; replace wk_net = (cnetpay / (52/9)) if cnetprd==6 & cnetpred==13 & znetpay==0; replace wk_net = (cnetpay / (52/10)) if cnetprd==6 & cnetpred==14 & znetpay==0; replace wk_net = (cnetpay / 13) if cnetprd==6 & cnetpred==15 & znetpay==0; replace wk_net = (cnetpay / 26) if cnetprd==6 & cnetpred==16 & znetpay==0; replace wk_net = (cnetpay * paidhrs) if cnetprd==6 & cnetpred==17 & zhours==0 & znetpay==0; replace wk_net = (cnetpay * days) if cnetprd==6 & cnetpred==18 & zhours==0 & znetpay==0; replace wk_net = 9999997 if cnetprd==6 & cnetpred>=19 & cnetpred~=. & znetpay==0; replace wk_net = 9999997 if cnetprd==6 & cnetpred==17 & zhours==1 & znetpay==0; replace wk_net = 9999997 if cnetprd==. & employee==1; replace wk_net = . if employee==0; lab var wk_net "weekly net pay"; /* missing net period */ gen znetprd = 1 if wk_net==9999997; replace znetprd = 0 if employee==1 & znetprd~=1; replace znetprd = . if employee==0; lab var znetprd "missing net period code"; replace wk_net = . if znetprd==1; /* annual net pay */ gen ann_net = (wk_net*52) if znetprd==0; replace ann_net=. if znetprd==1; lab var ann_net "annual net pay"; /* hourly net pay */ gen hr_net = (wk_net/ hours) if zhours==0 & znetprd==0; /* can get hourly pay even if hours missing if cnetpay is given hourly */


replace hr_net = cnetpay if cnetprd==6 & cnetpred==17 & zhours==1; lab var hr_net "hourly net pay"; /* monthly net pay */ gen mo_net = (wk_net * (13/3)) if znetprd==0; lab var mo_net "monthly net pay" ; end; donet; /******************************************** GROSS PAY VARIABLES *************************************/ /* this program converts gross pay into different periods depending on the period code */ cap pr drop dogross; pr def dogross; cap drop ann_gro; cap drop wk_gro; cap drop hr_gro; cap drop mo_gro; cap drop zgroprd; gen wk_gro = (cgropay) if cgroprd==1 & zgropay==0; replace wk_gro = (cgropay / 2) if cgroprd==2 & zgropay==0; replace wk_gro = (cgropay / 4) if cgroprd==3 & zgropay==0; replace wk_gro = (cgropay / (13/3)) if cgroprd==4 & zgropay==0; replace wk_gro = (cgropay / 52) if cgroprd==5 & zgropay==0; replace wk_gro = 9999997 if cgroprd==. & employee==1 & zgropay==0; replace wk_gro = (cgropay) if cgroprd==6 & cgropred==1 & zgropay==0; replace wk_gro = (cgropay / 2) if cgroprd==6 & cgropred==2 & zgropay==0; replace wk_gro = (cgropay / 3) if cgroprd==6 & cgropred==3 & zgropay==0; replace wk_gro = (cgropay / (13/3)) if cgroprd==6 & cgropred==4 & zgropay==0; replace wk_gro = (cgropay / 52) if cgroprd==6 & cgropred==5 & zgropay==0; replace wk_gro = (cgropay / 3) if cgroprd==6 & cgropred==6 & zgropay==0; replace wk_gro = (cgropay / 5) if cgroprd==6 & cgropred==7 & zgropay==0; replace wk_gro = (cgropay / 6) if cgroprd==6 & cgropred==8 & zgropay==0; replace wk_gro = (cgropay / 7) if cgroprd==6 & cgropred==9 & zgropay==0; replace wk_gro = (cgropay / 8) if cgroprd==6 & cgropred==10 & zgropay==0;


replace wk_gro = (cgropay / (26/3)) if cgroprd==6 & cgropred==11 & zgropay==0; replace wk_gro = (cgropay / (52/8)) if cgroprd==6 & cgropred==12 & zgropay==0; replace wk_gro = (cgropay / (52/9)) if cgroprd==6 & cgropred==13 & zgropay==0; replace wk_gro = (cgropay / (52/10)) if cgroprd==6 & cgropred==14 & zgropay==0; replace wk_gro = (cgropay / 13) if cgroprd==6 & cgropred==15 & zgropay==0; replace wk_gro = (cgropay / 26) if cgroprd==6 & cgropred==16 & zgropay==0; replace wk_gro = (cgropay * paidhrs) if cgroprd==6 & cgropred==17 & zhours==0 & zgropay==0; replace wk_gro = (cgropay * days) if cgroprd==6 & cgropred==18 & zhours==0 & zgropay==0; replace wk_gro = 9999997 if cgroprd==6 & cgropred>=19 & cgropred~=. & zgropay==0; replace wk_gro = 9999997 if cgroprd==9; /* one odd case here */ replace wk_gro = 9999997 if cgroprd==. & employee==1; replace wk_gro = . if employee==0; lab var wk_gro "weekly gross pay"; gen zgroprd = 1 if (wk_gro==9999997); replace zgroprd = 0 if (employee<=2 & zgroprd~=1); replace zgroprd = . if employee==0; lab var zgroprd "missing gross period code"; replace wk_gro = . if zgroprd==1; gen ann_gro = (wk_gro*52); lab var ann_gro "annual gross pay"; gen hr_gro = (wk_gro/ hours) if zhours==0; /* can get hourly pay even if hours missing if cgropay is given hourly */ replace hr_gro = cgropay if cgroprd==6 & cgropred==17 & zhours==1; lab var hr_gro "hourly gross pay"; gen mo_gro = (wk_gro * (13/3)); lab var mo_gro "monthly gross pay"; end; dogross; /*******************************FIXES TO PAY INFORMATION *************************************************/ /* store type of fix for reference */ /* Some are fixed because the ratio of net pay to gross pay is clearly wrong, and looking at the data it is obvious what the coding error has been.


Others are fixed because both gross and net pay are both clearly miscoded. Also some hours are miscoded */ cap drop fixtype; gen fixtype=0 if employee==1; lab var fixtype "type of fix to net and/ or gross pay"; lab def fixtype 0 "no fix" 1 "gro*10" 2 "net/100" 3 "net/10" 4 "net*100" 5 "net*10" 6 "net year to month" 7 "gro month to year" 8 "gro week to year" 9 "gro/10" 10 "gro week to month" 11 "individual fix" 12 "gro to missing" 13 "both/100" 14 "both year to month" 15 "both year to week" 16 "hours to missing" 17 "both month to year" 18 "both to missing" 19 "no fix- prop still odd" 20 "no fix- wages still odd", modify; lab val fixtype fixtype; /********* CHECK NET AGAINST GROSS FOR ODD AMOUNTS OR PERIOD CODES *******/ cap pr drop doprop; pr def doprop; cap drop hr_prop; gen hr_prop = hr_net/hr_gro; lab var hr_prop "ratio of net to gross hourly pay"; cap drop oddprop; gen oddprop =1 if (hr_prop<0.4|hr_prop>4) & hr_prop~=.; lab var oddprop "odd ratio of net to gross hourly pay"; cap log close; end; doprop; log using oddprop, replace; di "no fixes yet"; tab oddprop; log off; /******************** * FIX 1 * ********************/ /* lots of gross pay given annually needs to be multiplied by 10! */


replace fixtype=1 if (cgroprd==5|cgropred==5) & hr_prop>=5 & hr_prop~=. & cgropay<=6000; replace cgropay= cgropay*10 if fixtype==1; donet; dogross; doprop; log using oddprop, append; di "after fix 1:"; tab oddprop; log off; /******************** * FIX 2 * ********************/ /* there are a couple of monthly net pay which need to be divided by 100 */ replace fixtype=2 if (cnetprd==4|cnetpred==4) & hr_prop>=50 & hr_prop~=. & cnetpay>=80000; replace cnetpay = cnetpay/100 if fixtype==2; donet; dogross; doprop; log using oddprop, append; di "after fix 2:"; tab oddprop; log off; /******************** * FIX 3 * ********************/ /* there are some monthly net pay which need to be divided by 10 */ replace fixtype=3 if (cnetprd==4|cnetpred==4) & hr_prop>=5 & hr_prop~=. & cnetpay>=5000; replace cnetpay= cnetpay/10 if fixtype==3; donet; dogross; doprop; log using oddprop, append; di "after fix 3:"; tab oddprop; log off; /******************** * FIX 4 * ********************/ /* there are some net monthly and four weekly pay that need to be multiplied by 100 */ replace fixtype=4 if (cnetprd==4|cnetprd==3|cnetpred==4|cnetpred==3) & hr_prop<0.2 & cnetpay<50; replace cnetpay = cnetpay *100 if fixtype==4; donet; dogross;


doprop; log using oddprop, append; di "after fix 4:"; tab oddprop; log off; /******************** * FIX 5 * ********************/ /* there are some net monthly and four weekly pay that need mulitiplying by 10 */ replace fixtype=5 if (cnetprd==4|cnetprd==3|cnetpred==4|cnetpred==3) & hr_prop<0.2 & cnetpay<500; replace cnetpay = cnetpay *10 if fixtype==5 ; donet; dogross; doprop; log using oddprop, append; di "after fix 5:"; tab oddprop; log off; /******************** * FIX 6 * ********************/ /* lots of net monthly pay coded as yearly: period code needs to be changed */ replace fixtype=6 if (cnetprd==5|cnetpred==5) & cnetpay<=6000 & hr_prop<=0.1; replace cnetprd = 4 if fixtype==6; donet; dogross; doprop; log using oddprop, append; di "after fix 6:"; tab oddprop; log off; /******************** * FIX 7 * ********************/ /* there are some gross yearly pay coded as monthly and four weekly: period code needs to be changed */ replace fixtype = 7 if (cgroprd==3| cgroprd==4| cgropred==3| cgropred==4) & cgropay>=10000 & hr_prop<0.2; replace cgroprd = 5 if fixtype==7; donet; dogross; doprop; log using oddprop, append; di "after fix 7:"; tab oddprop;


log off; /******************** * FIX 8 * ********************/ /* there are some gross yearly pay coded as weekly: period code needs to be changed */ replace fixtype = 8 if (cgroprd==1|cgropred==1) & cgropay>=6000 & hr_prop<0.2; replace cgroprd = 5 if fixtype==8; donet; dogross; doprop; log using oddprop, append; di "after fix 8:"; tab oddprop; log off; /******************** * FIX 9 * ********************/ /* there are some gross yearly pay which need to be divided by 10 */ replace fixtype= 9 if (cgroprd==5|cgropred==5) & cgropay>=100000 & hr_prop<0.2; replace cgropay= cgropay/10 if fixtype==9; donet; dogross; doprop; log using oddprop, append; di "after fix 9:"; tab oddprop; log off; /******************** * FIX 10 * ********************/ /* there are some gross monthly pay coded as weekly: period code needs to be changed */ replace fixtype= 10 if (cgroprd==1 | cgropred==1) & cgropay>=350 & hr_prop<0.2; replace cgroprd= 4 if fixtype==10; donet; dogross; doprop; log using oddprop, append; di "after fix 10:"; tab oddprop; log off; /******************** * FIX 11 * ********************/ /* INDIVIDUAL FIXES - these come from eyeballing individual cases */


/* fixes to cgroprd */ replace cgroprd=1 if serial== 112006; replace cgroprd=4 if serial== 101716; replace cgroprd=4 if serial== 213363; replace cgroprd=6 if serial == 215256; /* net pay is 3 monthly and this looks more plausible */ replace cgropred=15 if serial== 215256; /* fixes to cnetprd */ replace cnetprd=4 if serial== 226017; replace cnetprd=4 if serial== 229217; replace cnetprd=1 if serial== 200306; /* from monthly to weekly, although makes net a bit high compared to gross */ replace cnetprd=1 if serial== 117261; /* from monthly to weekly, although makes net a bit high compared to gross */ replace cnetprd=1 if serial== 219899; /* from monthly to weekly, although makes net a bit high compared to gross */ replace cnetprd=1 if serial== 208399; /* from monthly to weekly, although makes net a bit high compared to gross */ replace cnetprd=1 if serial== 104694; /* from monthly to weekly, although makes net a bit high compared to gross */ replace cnetprd=1 if serial== 105206; /* from monthly to weekly, although makes net a bit high compared to gross */ replace cnetprd=2 if serial== 109860; /* gross pay is given fortnightly, and this looks more plausible */ /* fixes to cnetpay: make sure changes only get made to original miscoded value: ie. don't keep dividing or multiplying after the problem is already fixed */ replace cnetpay= onetpay/10 if serial== 124299; replace cnetpay= onetpay*10 if serial== 213092; replace cnetpay= onetpay*10 if serial== 229425; replace cnetpay=100 if serial== 119813; /* this one gets changed around by my different fixes, but eyeballing shows that it should be £100 */ replace cnetpay= onetpay*10 if serial== 111346; /* fixes to cgropay */ replace cgropay= ogropay*1000 if serial== 121730; replace cgropay= ogropay/10 if serial== 210049; replace cgropay= ogropay/10 if serial== 219518; donet; dogross; doprop; log using oddprop, append; di "after individual fixes "; tab oddprop; log off; /* update fixtype variable */ replace fixtype = 11 if serial== 112006; replace fixtype = 11 if serial== 101716; replace fixtype = 11 if serial== 213363; replace fixtype = 11 if serial== 215256;


replace fixtype = 11 if serial== 226017; replace fixtype = 11 if serial== 229217; replace fixtype = 11 if serial== 200306; replace fixtype = 11 if serial== 117261; replace fixtype = 11 if serial== 219899; replace fixtype = 11 if serial== 208399; replace fixtype = 11 if serial== 104694; replace fixtype = 11 if serial== 105206; replace fixtype = 11 if serial== 109860; replace fixtype = 11 if serial== 124299; replace fixtype = 11 if serial== 213092; replace fixtype = 11 if serial== 229425; replace fixtype = 11 if serial== 119813; replace fixtype = 11 if serial== 121730; replace fixtype = 11 if serial== 210049; replace fixtype = 11 if serial== 219518; replace fixtype = 11 if serial== 111346; /******************** * FIX 12 * ********************/ /* SET SOME GROSS TO MISSING if net looks sensisble, but gross looks rubbish, set gross to missing and then impute from net later */ replace cgropay=. if serial== 219802; replace cgropay=. if serial== 209108; replace cgropay=. if serial== 100831; /* update fixtype variable */ replace fixtype=12 if serial== 219802; replace fixtype=12 if serial== 209108; replace fixtype=12 if serial== 100831; /* update zgropay variable */ replace zgropay=1 if serial== 219802; replace zgropay=1 if serial== 209108; replace zgropay=1 if serial== 100831; donet; dogross; doprop; log using oddprop, append; di "after some set to missing "; tab oddprop; log off; /************************ ODD LOOKING WAGES --- WHERE GROSS AND NET ARE IN LINE BUT BOTH ARE WRONGLY CODED IN SOME WAY *******************/ so hr_gro;


cap log close; log using oddwage, replace; cap pr drop oddwage; pr def oddwage; cap drop oddwage; gen oddwage = 1 if (hr_gro>100 & hr_gro~=.) | hr_gro<2; replace oddwage= 0 if oddwage~=1; replace oddwage=. if employee==0; lab var oddwage "Strangely high or low gross hrly wage "; end; oddwage; /* record numbers that look too high or too low */ cap log close log using oddwage, append; di "NO FIXES YET "; tab oddwage; cap log close; /******************** * FIX 13 * ********************/ /*there are a number where the ratio of net to gross is fine, but both need to be divided by 100 */ /* Note these are guesses, but it would appear that full amounts including pence have been provided, but coded without the decimal point. In practice this applies to all gross hourly wages over £360 per hour ... */ replace fixtype =13 if hr_gro> 360 & hr_gro~=. ; replace cnetpay=onetpay/100 if fixtype==13; replace cgropay=ogropay/100 if fixtype==13; donet; dogross; oddwage; cap log close; log using oddwage, append; di "after FIX 13 "; tab oddwage; log off; /******************** * FIX 14 * ********************/ /* there are MANY where the ratio of net to gross is fine, but monthly pay has been recorded as annual. NOTE: A PLAUSIBLE ALTERNATIVE IS that both have lost a zero, and need to be multiplied by 10 (rather than 12 which is what we are implicitly doing here) Change it to monthly for now - can reconsider


Note: only change those whose implied hourly pay is less than £2 per hour, although there are many more who also look suspect Also only change to monthly those whose implied hourly pay is more than 36.7p per hour --- this is because from looking at them, the ones below this should be coded at WEEKLY instead */ /* quick data description of who these people are */ cap log close; log using oddwage, append; su hr_net hr_gro if cnetprd==5 & cgroprd==5 & cnetpay<2500 & cgropay<3000 & econact==1; su hr_net hr_gro if cnetprd==5 & cgroprd==5 & cnetpay<2500 & cgropay<3000 & econact==2; cap log close; replace fixtype = 14 if cnetprd==5 & cgroprd==5 & cnetpay<2500 & cgropay<3000 & hr_gro>0.367 & hr_gro<2; replace cnetprd=4 if fixtype ==14; replace cgroprd=4 if fixtype ==14; donet; dogross; oddwage; cap log close; log using oddwage, append; di "after FIX 14 "; tab oddwage; log off; /******************** * FIX 15 * ********************/ /* there are SOME where weekly pay has been recorded as annual. From looking at them, those whose implied hourly pay is less than 36.7p per hour ought to be coded as weekly */ replace fixtype = 15 if cnetprd==5 & cgroprd==5 & cnetpay<700 & cgropay<700 & hr_gro<0.367; replace cnetprd=1 if fixtype ==15; replace cgroprd=1 if fixtype ==15; donet; dogross; oddwage; cap log close; log using oddwage, append; di "after FIX 15 "; tab oddwage; log off;


/******************** * FIX 16 * ********************/ /* there are a number with IMPLAUSIBLE, VERY low weekly hours, which are not compatible with their self-reported work status, recorded in econact */ /* I'm not sure what to do with these? Set them to missing? Assume some average hours for them? Set to missing for now */ replace fixtype=16 if hours<=10 & hr_gro>50 & hr_gro~=. & econact==1; cap log close; log using oddwage, append; tab econact if hours<=10 & hr_gro>50 & hr_gro<=.; cap log close; replace hours=0 if fixtype==16; replace zhours=1 if fixtype==16; dohours; dogross; donet; oddwage; cap log close; log using oddwage, append; di "after FIX 16 "; tab oddwage; log off; /******************** * FIX 17 * ********************/ /* there are some who have annual amounts recorded as monthly: change period code to annual (* or else they could need to be divided by 10? see above? *) */ replace fixtype = 17 if (cnetprd==3| cnetprd==4) & cnetpay>10000 & hr_net>100 & hr_net~=.; replace cnetprd = 5 if (cnetprd==3| cnetprd==4) & cnetpay>10000 & hr_net>100 & hr_net~=.; replace fixtype = 17 if (cgroprd==3| cgroprd==4) & cgropay>10000 & hr_gro>100 & hr_gro~=.; replace cgroprd = 5 if (cgroprd==3| cgroprd==4) & cgropay>10000 & hr_gro>100 & hr_gro~=.; dogross; donet; oddwage; cap log close; log using oddwage, append; di "after FIX 17 "; tab oddwage; log off; /********************


* FIX 18 * ********************/ /* there are 8 obs where cgropay and cnetpay both coded as 1 : clearly nothing to do with their pay, although (except for 1 obs who "refused" to give period codes), period codes and hours work information is provided. Set to missing for now: can good data for these be recovered? */ replace fixtype=18 if cnetpay==1 & cgropay==1 & hr_gro~=.; replace cnetpay=. if fixtype==18; replace znetpay=1 if fixtype==18; replace cgropay=. if fixtype==18; replace zgropay=1 if fixtype==18; donet; dogross; oddwage; cap log close; log using oddwage, append; di "after FIX 18 "; tab oddwage; log off; /******************** * FIX 11: AGAIN * ********************/ /* some more individual fixes to gross and net pay */ /* WHAT TO DO?? 1) Need to find a fix for serial 122904 : 10 net pay, 15 gross pay --- multiply by 100? already has had net period code changed from annual to monthly to be more in line with the gross period code, but this may not be the right fix in this case? 2) SERIAL 109130: monthly gross and net pay given as £55 for 56 hours per week = 22p per hour 3) SERIAL 218767, 208683,230872, 20174: really high annual pay - see below for comparison to wave 5 info */ li serial chrtid rw* wk_net _m if serial==218767; li serial chrtid rw* wk_gro _m if serial==228991; li serial chrtid rw* wk_net _m if serial==208683; li serial chrtid rw* wk_net _m if serial==230872; li serial chrtid rw* wk_net _m if serial==201714; replace cgropay=ogropay/10 if serial ==228991; /* based on weekly net pay in wave 5 of £605.305 -- more plausible that this one is in the region of £600 not £6000 */ replace fixtype=11 if serial==228991; drop if _m==2; donet; dogross; oddwage;


cap log close; log using oddwage, append; di "after more indiviual fixes:"; tab oddwage; log off; /******************** * FIXES 19 AND 20: NOT REALLY A FIX - * ********************/ /* if proportion of net to gross is still odd, but no fix has been made */ replace fixtype=19 if oddprop==1 & fixtype==0; /* if gross hourly wage is still > £100 or < £2, but no fix has been made */ replace fixtype=20 if oddwage==1 & fixtype==0; /****************************************** SET UP SOME CATEGORY VARIABLES FOR REFERENCE ******************/ /* Missing status variables for cnetpay and cgropay: for each net pay obs missing: some can be imputed from gross, for others, net and gross are both missing. */ /* both gross and net missing */ cap drop bothmiss; gen bothmiss =1 if znetpay==1 & zgropay==1; replace bothmiss =0 if bothmiss~=1; replace bothmiss = . if employee==0; lab var bothmiss "both wages missing"; /* can impute net from gross */ cap drop impnet; gen impnet =1 if znetpay==1 & zgropay~=1; replace impnet =0 if impnet~=1; replace impnet =. if employee==0; lab var impnet "can impute net from gross"; /* can impute gross from net */ cap drop impgro; gen impgro =1 if zgropay==1 & znetpay~=1; replace impgro =0 if impgro~=1; replace impgro =. if employee==0; lab var impgro "can impute gross from net"; /* how to fix missing net pay */ cap drop fixnet; gen fixnet =1 if impnet==1; replace fixnet =2 if bothmiss==1; replace fixnet =0 if znetpay==0; replace fixnet =. if employee==0; lab var fixnet "how to fix missing net pay";


lab def fixnet 0 "no fix needed" 1 "can impute from gross" 2 "set to missing", modify; lab val fixnet fixnet; /*how to fix missing gross pay */ cap drop fixgro; gen fixgro =1 if impgro==1; replace fixgro =2 if bothmiss==1; replace fixgro =0 if zgropay==0; replace fixgro =. if employee==0; lab var fixgro "how to fix missing gross wage"; lab def fixgro 0 "no fix needed" 1 "can impute from net" 2 "set to missing", modify; lab val fixgro fixgro; replace fixtype=. if bothmiss==1; /*********************************************REASONS WHY NO HOURLY PAY AVAILABLE ***************************/ cap drop zhr_net; lab def zhr_net 0 "not missing" 1 "imputed from gross" 2 "missing pay: cannot impute" 3 "missing period code: cannot impute" 4 "missing hours, no hourly pay given" , modify; gen zhr_net = 1 if fixnet==1; replace zhr_net = 1 if fixnet==0 & znetprd==1 & zgropay==0 & zgroprd==0; /* net period code is missing */ replace zhr_net = 2 if fixnet==2; replace zhr_net = 3 if (fixnet==0| fixnet==1) & znetprd==1 & zgroprd==1; replace zhr_net = 4 if (fixnet==0 & znetprd==0 & zhours==1)| (fixnet==1 & zgroprd==0 & zhours==1); replace zhr_net = 0 if employee==1 & zhr_net==.; /* set back to not missing or can impute if pay is given hourly, or could impute from gross pay given hourly */ replace zhr_net = 0 if cnetprd==6 & cnetpred==17 & zhours==1; replace zhr_net = 1 if zhr_net==5 & fixnet==1 & cnetprd==6 & cnetpred==17 & zhours==1; replace zhr_net = . if employee==0; lab val zhr_net zhr_net; lab var zhr_net "missing hourly net pay"; cap drop zhr_gro; gen zhr_gro = 1 if fixgro==1; replace zhr_gro = 1 if fixgro==0 & zgroprd==1 & znetpay==0 & znetprd==0; /* gross period code is missing */ replace zhr_gro = 2 if fixgro==2; replace zhr_gro = 3 if zgroprd== 1 & znetprd ==1 & (fixgro==0 | fixgro==1); replace zhr_gro = 4 if (fixgro==0 & zgroprd==0 & zhours==1)| (fixgro==1 & znetprd==0 & zhours==1); replace zhr_gro = 0 if employee==1 & zhr_gro==.;


/* set back to not missing or can impute if pay is given hourly, or could impute from gross pay given hourly */ replace zhr_gro = 0 if cgroprd==6 & cgropred==17 & zhours==1; replace zhr_net = 1 if zhr_gro==5 & fixgro==1 & cnetprd==6 & cnetpred==17 & zhours==1; replace zhr_gro = . if employee==0; lab def zhr_gro 0 "not missing" 1 "imputed from net" 2 "missing pay: cannot impute" 3 "missing period code: cannot impute" 4 "missing hours, no hourly pay given" , modify; lab val zhr_gro zhr_gro; lab var zhr_gro "missing hourly gro pay"; /****************************** do the imputations *****************************************/ do "e:\2000cohorts\do files\impute"; replace hr_gro= Ihr_gro if zhr_gro==1; replace wk_gro= Iwk_gro if zhr_gro==1; replace mo_gro= Imo_gro if zhr_gro==1; replace ann_gro= Iann_gro if zhr_gro==1; replace hr_net= Ihr_net if zhr_net==1; replace wk_net= Iwk_net if zhr_net==1; replace mo_net= Imo_net if zhr_net==1; replace ann_net= Iann_net if zhr_net==1; /****************************** inflate to Jan 01 ****************************/ /* these will have been merged in from Iintdate.dta */ cap drop rhr_gro; gen rhr_gro = hr_gro/ rpi_int; lab var rhr_gro "real gross hourly earnings, j01 prices"; cap drop lhw; gen lhw = log(hr_gro); lab var lhw "log real gross hourly earnings, j01 prices"; /************************** code up some simple characteristics *****************/ cap drop male; gen male =1 if cmsex==1; replace male = 0 if cmsex==2; lab var male "male dummy"; cap drop sampsex; gen sampsex =1 if sample==1 & male==1; replace sampsex =2 if sample==2 & male==1; replace sampsex =3 if sample==1 & male==0; replace sampsex =4 if sample==2 & male==0; lab def sampsex 1 "bcs70 man" 2 "ncds man" 3 "bcs70 woman" 4 "ncds woman", modify; lab val sampsex sampsex; cap drop bcsman; cap drop ncdsman;


cap drop bcswom; cap drop ncdswom; gen bcsman = 1 if sampse==1; gen ncdsman = 1 if sampse==2; gen bcswom = 1 if sampse==3; gen ncdswom = 1 if sampse==4; replace bcsman = 0 if bcsman~=1; replace ncdsman = 0 if ncdsman~=1; replace bcswom = 0 if bcswom~=1; replace ncdswom = 0 if ncdswom~=1; /* log some results */ cap log close; log using misspay, replace; /* BASIC PAY VARIABLES */ tab fixnet sample, col freq; tab fixgro sample, col freq; /* HOURS */ tab zhours sample, col freq; /* HOURLY PAY VARIABLES */ tab zhr_net sample, col freq; tab zhr_gro sample, col freq; tab zhr_net econact, col freq; tab zhr_gro econact, col freq; tab zhr_net cmsex, col freq; tab zhr_gro cmsex, col freq; /* FIXES TO PAY VARIABLES */ tab fixtype sample, col freq; /**************** hourly gross pay descriptives **************************/ su hr_gro, det; table sampsex, c(median hr_gro); table sampsex, c(mean hr_gro); table sample econact, c(median hr_gro), if employee==1; table sample econact, c(mean hr_gro), if employee==1; table sample econact, c(median hr_gro), if cmsex==1 & employee==1; table sample econact, c(mean hr_gro), if cmsex==2 & employee==1; gr hr_gro if hr_gro<40 & oddwage==0 & oddprop==0, bin(50) xlab(0,3,10,20,40) by(sampsex) ylab(0,0.04,0.08,0.12,0.16); reg lhw ncdsman bcsman ncdswom ; log close; /******************************FINALLY, SET NEW IMPUTED PAY AND PERIOD VARIABLES, AND RETURN CNETPAY, ETC. TO THEIR ORIGINALS ******************************/


cap drop Inetpay; gen Inetpay=cnetpay; lab var Inetpay "imputed/corrected net pay variable"; cap drop Igropay; gen Igropay=cgropay; lab var Igropay "imputed/corrected gross pay variable"; cap drop Inetprd; gen Inetprd=cnetprd; lab var Inetprd "imputed/corrected net period variable"; label values Inetprd cnetprd; cap drop Igroprd; gen Igroprd=cgroprd; lab var Igroprd "imputed/corrected gross period variable"; label values Igroprd cgroprd; /* return cnet and gro variables to their originals */ replace cnetpay=onetpay; replace cnetprd=onetprd; replace cgropay=ogropay; replace cgroprd=ogroprd;


IMPUTE.DO /********************************************IMPUTATION OF NET AND GROSS PAY *******************************/ /* assume all are contracted IN to SERPs */ /* assume none get WFTC through pay packet */ # delimit; set more 1; /******** interview dates and tax year **********************************************************/ /* use corrected intdate variable once have got this together properly..*/ cap drop intmth; cap drop intyr; cap drop temp; gen str2 temp= substr(intdate,3,2); gen intmth= real(temp); cap drop temp; gen str4 temp = substr(intdate,5,4); gen intyr= real(temp); cap drop taxyear; gen taxyear=1999 if ((intyr==1999)| (intyr==2000 & intmth<4)); replace taxyear=2000 if taxyear~=1999; lab var taxyear "tax year"; /**** Because there was still the MCA in 1999, need to check if there are any children ***********/ /* sum across household members to see if any are children u19 */ cap pr drop makekid; pr def makekid; local loop = 2; while `loop' <10 {; /* include foster kids for now */ replace kidinhh = kidinhh+ 1 if reltoke`loop' ==3 & age`loop'<17 | reltoke`loop' ==4 & age`loop'<17 | reltoke`loop' ==5 & age`loop'<17 | reltoke`loop' ==6 & age`loop'<17 | reltoke`loop' ==7 & age`loop'<17; replace maybek = maybek+1 if (reltoke`loop' ==3 & (age`loop'==18|age`loop'==19| age`loop'==999)) | (reltoke`loop' ==4 & (age`loop'==18|age`loop'==19| age`loop'==999)) | (reltoke`loop' ==5 & (age`loop'==18|age`loop'==19| age`loop'==999)) | (reltoke`loop' ==6 & (age`loop'==18|age`loop'==19| age`loop'==999)) | (reltoke`loop' ==7 & (age`loop'==18|age`loop'==19| age`loop'==999)); local loop = `loop' +1;


}; end; /*********** need to know marital status and if there are children in the household for MCA in 1999 */ cap drop kidinhh; gen kidinhh=0; lab var kidinhh "kid u17 in household grid"; cap drop maybek; gen maybek=0; lab var maybek "kid 18,19, missing age in household grid"; makekid; cap drop mca; gen mca=1 if taxyear==1999 & ms==1; replace mca = 1 if taxyear==1999 & (kidinhh>0|maybek>0); replace mca =0 if mca ~=1; lab var mca "getting married couples allowance?"; /**************************** some helpful indicators ******************************************************/ /* can impute net from gross */ cap drop impnet; gen impnet =1 if znetpay==1 & zgropay~=1; replace impnet =0 if impnet~=1; replace impnet =. if employee==0; lab var impnet "can impute net from gross"; /* can impute gross from net */ cap drop impgro; gen impgro =1 if zgropay==1 & znetpay~=1; replace impgro =0 if impgro~=1; replace impgro =. if employee==0; lab var impgro "can impute gross from net"; /*********************** some parameters of the tax system *******************************************/ cap drop lel; cap drop uel; cap drop pa; cap drop thresh1; cap drop thresh2; cap drop lowrate; cap drop baserate; cap drop highrate; gen lel=67 if taxyear==2000; replace lel=66 if taxyear==1999; lab var lel "national insurance lower earnings limit"; gen uel =535 if taxyear==2000;


replace uel=500 if taxyear==1999; lab var uel "national insurance upper earnings limit"; gen pa = 4385 if taxyear==2000; replace pa = 4335 if taxyear==1999; lab var pa "personal allowance"; gen thresh1 = 1520 if taxyear==2000; replace thresh1 = 1500 if taxyear==1999; lab var thresh1 "lower rate limit"; gen thresh2 = 28400 if taxyear==2000; replace thresh2 = 28000 if taxyear==1999; lab var thresh2 "basic rate limit"; gen lowrate= 0.1; lab var lowrate "lower rate"; gen baserate = 0.22 if taxyear==2000; replace baserate = 0.23 if taxyear==1999; lab var baserate "basic rate"; gen highrate = 0.4; lab var highrate "higher rate"; cap drop notax; cap drop lowband; cap drop baseband; cap drop highband; /* put people in income tax bands according to their net incomes */ gen notax =1 if (ann_net<=pa); gen lowband =1 if (ann_net>pa) & (ann_net <= (pa + (0.9 * thresh1))); gen baseband=1 if (ann_net > (pa + (0.9 * thresh1))) & (ann_net <= (pa + (0.9 * thresh1) + ((1-baserate) * (thresh2-thresh1)))); gen highband=1 if (ann_net >=(pa + (0.9 * thresh1) + ((1-baserate) * (thresh2-thresh1)))); /*************************************** NATIONAL INSURANCE *****************************************/ /* FROM GROSS */ cap drop wk_nig; gen wk_nig = 0 if wk_gro<lel ; replace wk_nig = 0.1* (wk_gro-lel) if wk_gro>=lel & wk_gro<=uel ; replace wk_nig = 0.1* (uel-lel) if wk_gro>uel ; lab var wk_nig "from gross: imputed weekly NI payments"; /* FROM NET */ cap drop wk_nin;


gen wk_nin = 0 if wk_net<=lel ; replace wk_nin = ((wk_net-lel)/9) if wk_net>lel & wk_net<= uel*0.9 ; replace wk_nin = (0.1* (uel- lel)) if wk_net>uel*0.9 ; lab var wk_nin "from net: imputed weekly NI payments"; /************************************** INCOME TAX *****************************************************/ /* tax is worked out separately for those whose income falls: 1. below personal allowance 2. in 10p band 3. in basic rate band 4. in higher rate band */ /* FROM GROSS */ cap drop taxann; gen taxann = ann_gro-pa; cap drop ann_itg; gen ann_itg =0 if taxann <=0 ; replace ann_itg = lowrate * taxann if taxann > 0 & (taxann < thresh1) ; replace ann_itg =(lowrate * (thresh1)) + (baserate * (taxann - thresh1)) if (taxann>= thresh1) & ( taxann< thresh2) ; replace ann_itg =(lowrate * (thresh1)) + (baserate * (thresh2 - thresh1)) + (highrate * (taxann - thresh2)) if (taxann>=thresh2) ; /* sort out the MCA here: GIVE them the WHOLE mca for now - ie don't try to assume split between partner or anything! */ replace ann_itg = ann_itg - 197 if mca==1; replace ann_itg = 0 if ann_itg <0; lab var ann_itg "from GROSS: imputed annual income tax payments"; /* FROM NET */ cap drop ann_itn; gen ann_itn =0 if notax==1; replace ann_itn = (lowrate/(1-lowrate)) *(ann_net-pa) if lowband==1; replace ann_itn = ((baserate/(1-baserate))*(ann_net-pa)) - (((baserate-lowrate) /(1-baserate)) * thresh1) if baseband==1; replace ann_itn = ((highrate/(1-highrate))*(ann_net-pa)) - (((highrate-baserate)/(1-highrate)) * thresh2) - (((baserate-lowrate)/(1-highrate)) * thresh1) if highband==1; replace ann_itn = ann_itn -197 if mca==1; replace ann_itn = 0 if ann_itn<0; lab var ann_itn "from NET: imputed annual income tax payments";


cap drop Iann_gro; cap drop Iann_net; gen Iann_gro = ann_net + (wk_nin*52) + ann_itn ; lab var Iann_gro " gross annual income imputed from net "; gen Iann_net = ann_gro - (wk_nig*52) - ann_itg ; lab var Iann_net " net annual income imputed from gross "; /* need to translate these back into other periods : hourly, weekly */ /* weekly */ cap drop Iwk_gro; cap drop Iwk_net; gen Iwk_gro= Iann_gro/52; lab var Iwk_gro "weekly gross imputed from net"; gen Iwk_net= Iann_net/52; lab var Iwk_net "weekly net imputed from gross"; /*monthly */ cap drop Imo_gro; cap drop Imo_net; gen Imo_gro= Iann_gro/12; lab var Imo_gro "monthly gross imputed from net"; gen Imo_net = Iann_net/12; lab var Imo_net "monthly net imputed from gross"; /* hourly */ cap drop Ihr_gro; cap drop Ihr_net; gen Ihr_gro = Iwk_gro/hours; lab var Ihr_gro "hourly gross imputed from net"; gen Ihr_net = Iwk_net/hours; lab var Ihr_net "hourly net imputed from gross"; /***********************log some results **********************************************/ cap drop prop; gen prop = hr_net/hr_gro; replace prop = hr_net/Ihr_gro if zhr_gro==1; replace prop = Ihr_net/hr_gro if zhr_net==1; cap drop imputax*; gen imputax1 = Ihr_net/hr_gro; gen imputax2 = hr_net/Ihr_gro; lab var imputax1 "ratio: imputed net given gross"; lab var imputax2 "ratio: imputed gross given net"; cap drop reported; gen reported = hr_net/hr_gro; cap log close; log using imputation, replace; su prop;


su prop if zhr_net~=1 & zhr_gro~=1, det; su prop if zhr_net==1,det; su prop if zhr_gro==1,det; su hr_gro Ihr_gro hr_net Ihr_net; sum imputax1 imputax2 reported; cap log close;


PARTNER.DO /* this sorts out partner's income 1. set up missing variable 2. put into weekly/ annual 3. any fixes? 4. impute gross from net */ #delimit; set more 1; /************* set original in stone **************************************/ cap drop opnet; gen opnet = pnetpay; lab var opnet "original pnetpay"; cap drop opnetprd; gen opnetprd = pnetprd; lab var opnetprd "original pnetprd"; cap drop opnpred; gen opnpred = pnetpred; lab var opnpred "original pnetpred"; /**** EARNINGS STATUS : this is different from the cm coding, as there are earnings figs for people across all peconact, not just employees *****************/ cap drop pearns; gen pearns =1 if pnetpay~=.; replace pearns =0 if pnetpay==.; lab var pearns "partner's earnings reported"; /***************** missing partner's pay ************************************/ cap drop zpnetpay; gen zpnetpay =1 if opnet== 99999998|opnet==99999999|opnet==1; replace zpnetpay =0 if zpnetpay~=1; replace zpnetpay =. if pnetpay==.; lab var zpnetpay "missing pnetpay"; /******************* period codes ************************************************/ /*the coding frame for pnetprd: 1 one week 2 a fortnight 3 four weeks 4 a calendar month 5 a year or 6 other period the coding frame for pnetpred: 1 one week 2 a fortnight/2 weeks 3 four weeks 4 a calendar month 5 a year (12 months/52 weeks)


6 three weeks 7 five weeks 8 six weeks 9 seven weeks 10 eight weeks 11 two calendar months 12 eight times a year 13 nine times a year 14 ten times a year 15 three months/13 weeks 16 six months/26 weeks 17 hourly 18 daily 19 one off lump sum 20 some other period 21 varies 22 refused 23 some other comment 24 irrelevant/unspecific response */ /*********************************************** NET PAY VARIABLES **************************************/ /* First make sure not to start manipulating missing values 9999997 and 9999998 ! */ replace pnetpay =. if zpnetpay==1; /* need to guess days per week (5 give their pay daily) hours per week (3 give their pay hourly) assume the number of hours (and days based on this) is the median employee hours of the cohort members-- this is 20 for part time males and females, 40 for full-time females, 45 for full-time males (assume partner's sex is opp of cm sex -- too much effort to get proper code from household grid for now -- out of these 8 its pretty much certain to be true!) */ cap drop phours; gen phours = 20 if (peconact==2|peconact==4); replace phours = 40 if (peconact==1|peconact==3) & cmsex==1; replace phours = 45 if (peconact==1|peconact==3) & cmsex==2; lab var phours "partner's hours: estimated"; cap drop pdays; gen pdays = 2.5 if (peconact==2|peconact==4); replace pdays = 5 if (peconact==1|peconact==3); lab var pdays "partner's days: estimated"; /* this program converts net pay into different periods depending on the period code */ cap pr drop dopnet; pr def dopnet;


cap drop ann_pnet; cap drop wk_pnet; cap drop zpnetprd; gen wk_pnet = (pnetpay) if pnetprd==1 & zpnetpay==0; replace wk_pnet = (pnetpay / 2) if pnetprd==2 & zpnetpay==0; replace wk_pnet = (pnetpay / 4) if pnetprd==3 & zpnetpay==0; replace wk_pnet = (pnetpay / (13/3)) if pnetprd==4 & zpnetpay==0; replace wk_pnet = (pnetpay / 52) if pnetprd==5 & zpnetpay==0; replace wk_pnet = 9999997 if pnetprd==. & pearns==1 & zpnetpay==0; replace wk_pnet = (pnetpay) if pnetprd==6 & pnetpred==1 & zpnetpay==0; replace wk_pnet = (pnetpay / 2) if pnetprd==6 & pnetpred==2 & zpnetpay==0; replace wk_pnet = (pnetpay / 3) if pnetprd==6 & pnetpred==3 & zpnetpay==0; replace wk_pnet = (pnetpay / (13/3)) if pnetprd==6 & pnetpred==4 & zpnetpay==0; replace wk_pnet = (pnetpay / 52) if pnetprd==6 & pnetpred==5 & zpnetpay==0; replace wk_pnet = (pnetpay / 3) if pnetprd==6 & pnetpred==6 & zpnetpay==0; replace wk_pnet = (pnetpay / 5) if pnetprd==6 & pnetpred==7 & zpnetpay==0; replace wk_pnet = (pnetpay / 6) if pnetprd==6 & pnetpred==8 & zpnetpay==0; replace wk_pnet = (pnetpay / 7) if pnetprd==6 & pnetpred==9 & zpnetpay==0; replace wk_pnet = (pnetpay / 8) if pnetprd==6 & pnetpred==10 & zpnetpay==0; replace wk_pnet = (pnetpay / (26/3)) if pnetprd==6 & pnetpred==11 & zpnetpay==0; replace wk_pnet = (pnetpay / (52/8)) if pnetprd==6 & pnetpred==12 & zpnetpay==0; replace wk_pnet = (pnetpay / (52/9)) if pnetprd==6 & pnetpred==13 & zpnetpay==0; replace wk_pnet = (pnetpay / (52/10)) if pnetprd==6 & pnetpred==14 & zpnetpay==0; replace wk_pnet = (pnetpay / 13) if pnetprd==6 & pnetpred==15 & zpnetpay==0; replace wk_pnet = (pnetpay / 26) if pnetprd==6 & pnetpred==16 & zpnetpay==0; replace wk_pnet = (pnetpay * phours) if pnetprd==6 & pnetpred==17 & zpnetpay==0; replace wk_pnet = (pnetpay * pdays) if pnetprd==6 & pnetpred==18 & zpnetpay==0; replace wk_pnet = 9999997 if pnetprd==6 & pnetpred>=19 & pnetpred~=. & zpnetpay==0; replace wk_pnet = 9999997 if pnetprd==. & pearns==1;


lab var wk_pnet "partner weekly net pay"; /* missing net period */ gen zpnetprd = 1 if wk_pnet==9999997; replace zpnetprd = 0 if pearns==1 & zpnetprd~=1; replace zpnetprd = . if pearns==0; lab var zpnetprd "missing net period code"; replace wk_pnet = . if zpnetprd==1; /* annual net pay */ gen ann_pnet = (wk_pnet*52) if zpnetprd==0; replace ann_pnet=. if zpnetprd==1; lab var ann_pnet "partner annual net pay"; /* ROUGH guess at hourly net pay -- by dividing by estimated hours */ cap drop hr_pnet; gen hr_pnet = wk_pnet/phours; replace hr_pnet =. if zpnetprd==1; end; dopnet; cap pr drop podd; pr def podd; cap drop podd; gen podd=1 if (hr_pnet<2 | hr_pnet>100) & ann_pnet~=.; lab var podd "odd partner's wage"; tab podd; end; podd; /******************** FIXES TO PARTNER'S PAY ***************/ /* FIX ONE A LOAD CODED AS WEEKLY WHICH I AM CHANGING TO ANNUAL ---- SOME COULD HAVE THE DECIMAL POINT IN THE WRONG PLACE, ITS IMPOSSIBLE TO KNOW WHICH, BUT IT MATTERS SINCE WHAT I AM DOING DIVIDES BY 52, WHEREAS THE OTHER DIVIDES BY 100!!!! */ replace pnetprd=5 if (serial ==111207 | serial == 205291 | serial == 100875 | serial == 127091 | serial ==107752 | serial == 108363 | serial == 211375); /* FIX TWO A LOAD CODED AS ANNUAL WHICH MUST BE MONTHLY */ replace pnetprd = 4 if hr_pnet<1 & pnetpay>500 & pnetprd==5; /* FIX 3: these need to be multiplied by 1000 I think ! */ replace pnetpay=opnet*1000 if serial==231141|serial==220423|serial==218974; /* A LOAD CODED AS ANNUAL WHICH I THINK ARE WEEKLY - ALTHOUGH THEY COULD BE HOURLY AND WITHOUT A DECIMAL POINT...! */


replace pnetprd=1 if serial==223385|serial==208807|serial==128051|serial==230870|serial==208819|serial==223287; /* FIX 4: these are coded as monthly which should be annual */ replace pnetprd = 5 if (serial == 129075|serial==226193|serial==200516|serial==208649); su pnetprd; /* FIX 5 : decimal point has gone missing: divide by 100 */ replace pnetpay = opnet /100 if serial==109177|serial==228795; /* FIX 6: some individual fixes */ replace pnetpay =1100 if serial==129849; /* Fix 7: these ought to be monthly ! */ replace pnetprd = 4 if serial==230227|serial==111346| serial==228451|serial==206491|serial==108097|serial==220322| serial==110103|serial==227417|serial==222487|serial==130222| serial==218190|serial==112998|serial==209574|serial==105600| serial==222518|serial==123032|serial==229425|serial==120296; /* these need to be *100 */ replace pnetpay = opnet *100 if serial== 216308|serial==201426|serial==111346|serial==222511; /* these need to be * 10 */ replace pnetpay= opnet * 10 if serial==113089|serial==214352|serial==211187|serial==104141| serial==209738|serial==219897|serial==228918|serial==111609| serial==222537|serial==104729|serial==125905|serial==111127| serial==217108|serial==111811|serial==212263; dopnet; replace pnetpay = pnetpay*10 if hr_pnet>=0.5 & hr_pnet<=1; /* these look like hourly */ replace pnetprd=6 if serial==225242|serial==105684|serial==200634|serial==112685; replace pnetpred=17 if serial==225242|serial==105684|serial==200634|serial==112685; /* change to missing - can't really tell what they should be...*/ replace pnetpay=. if serial==113463|serial==222918|serial==219680|serial==117299|serial==130637| serial==202656|serial==219418|serial==124541|serial==127480|serial==214513; /* these look like weekly */ replace pnetprd=1 if pnetprd==4 & hr_pnet>=1.02 & hr_pnet<2; dopnet; podd;


/* lots more needed, but give up for now.... / JUST IMPUTE A GROSS FIGURE FROM THIS NET -- athough the net is given net of pension contributions, so gross will not be accurate! */ /*************************************** NATIONAL INSURANCE *****************************************/ /* FROM NET */ cap drop wk_nip; gen wk_nip = 0 if wk_pnet<=lel ; replace wk_nip = ((wk_pnet-lel)/9) if wk_pnet>lel & wk_pnet<= uel*0.9 ; replace wk_nip = (0.1* (uel- lel)) if wk_pnet>uel*0.9 ; lab var wk_nip "from net: partner's imputed weekly NI payments"; /************************************** INCOME TAX *****************************************************/ /* tax is worked out separately for those whose income falls: 1. below personal allowance 2. in 10p band 3. in basic rate band 4. in higher rate band */ cap drop pnotax; cap drop plowb; cap drop pbaseb; cap drop phighb; /* put partners in income tax bands according to their net incomes */ gen pnotax =1 if (ann_pnet<=pa); gen plowb =1 if (ann_pnet>pa) & (ann_pnet <= (pa + (0.9 * thresh1))); gen pbaseb=1 if (ann_pnet > (pa + (0.9 * thresh1))) & (ann_pnet <= (pa + (0.9 * thresh1) + ((1-baserate) * (thresh2-thresh1)))); gen phighb=1 if (ann_pnet >=(pa + (0.9 * thresh1) + ((1-baserate) * (thresh2-thresh1)))); /* FROM NET */ cap drop ann_itp; gen ann_itp =0 if notax==1; replace ann_itp = (lowrate/(1-lowrate)) *(ann_pnet-pa) if lowband==1; replace ann_itp = ((baserate/(1-baserate))*(ann_pnet-pa)) - (((baserate-lowrate) /(1-baserate)) * thresh1) if baseband==1; replace ann_itp = ((highrate/(1-highrate))*(ann_pnet-pa)) - (((highrate-baserate)/(1-highrate)) * thresh2) - (((baserate-lowrate)/(1-highrate)) * thresh1) if highband==1;


/* don't know if its right to give the whole MCA to the partner -- when imputing before I gave it all to the cohort memeber think about doing this properly later. */ replace ann_itp = ann_itp -197 if mca==1; replace ann_itp = 0 if ann_itp<0; lab var ann_itp "from NET: partner's imputed annual income tax payments"; cap drop ann_pgro; gen ann_pgro = ann_pnet + (wk_nip*52) + ann_itp ; lab var ann_pgro "partner's gross annual income imputed from net "; /* need to translate these back into other periods : hourly, weekly */ /* weekly */ cap drop wk_pgro; gen wk_pgro= ann_pgro/52; lab var wk_pgro "partner's weekly gross imputed from net"; /* hourly */ cap drop hr_pgro; gen hr_pgro = wk_pgro/phours; lab var hr_pgro "partner's hourly gross imputed from net"; /************* put originals back **************************************/ #delimit; cap drop Ipnetpay; cap drop Ipnetprd; cap drop Ipnpred; rename pnetpay Ipnetpay; rename pnetprd Ipnetprd; rename pnetpred Ipnpred; rename opnet pnetpay; rename opnetprd pnetprd; rename opnpred pnetpred; label values pnetprd pnetprd; label values pnetpred pnetpred;


APPENDIX 5: Basic Skills and other variables

Documentation Samantha Parsons of the Centre for Longitudinal Studiesrelating to basic skills and other varibels is reproduced below.


([email protected])

PRELIMINARY ANALYSIS OF NCDS/BCS70 NEW 1999/2000 SURVEY DATA

i. Age CM and CM partner left full-time education ii. Maths difficulties iii. Reading difficulties iv. Writing difficulties v. Other difficulties of having 3R problem vi. Skills questions vii. Course attendance viii. Computer use at home and work ix. Illegal drug use x. Diet, food and exercise xi. Eating problems xii. Smoking xiii. Drinking xiv. Accidents


AGE LEFT FULL-TIME EDUCATION QUESTIONS variable system-missing user-missing: 8 (^



n % n % n % actagel2 607 2.7 ^1 0 22072 actagel2 details age CM left full-time education agelfte2 66 0.3 6 0 36 0.2 22572 agelfte2 captures CMs who were in full-time education at time of interview furthed2 607 2.7 22073 furthed2 details CMs who returned to full-time education less than 3 years after initially leaving full-time education lftmore2 20522 90.5 ^2 2156 lftmore2 details age CM left second period of full-time education plefted 5981 26.4 ^12 0.1 ^347 2.1 16340 plefted details age CM partner left full-time education Derived AGE LEFT FULL-TIME EDUCATION Variables



valid n

n % n % n % Ageftedu 109 0.5 22571 6 category variables combining information from actagel2 and agelfte2. Values: 1'before 16' 2'at 16' 3'post 16' 4'post 18' 5'post 21' 6'still in ft edu'. Agelfted 109 0.5 22571 continuous variable combining information from actagel2, agelfte2, furthed2 and lftmore2. Value range: 14 to 42, 99=still in education Ageledu 109 0.5 22571 6 category variable collapsing information in agelfted. Values: 1'before 16' 2'at 16' 3'post 16' 4'post 18' 5'post 21' 6'still in ft edu' plefted1 6340 28.0 16340 5 category variable collapsing information in plefted. Values: 1'before 16' 2'at 16' 3'post 16' 4'post 18' 5'post 21'


SPSS SYNTAX age CM left full-time education frequencies variables=actagel2 agelfte2 furthed2 lftmore2. missing values actagel2 lftmore2 (99). missing values agelfte2 (8,9). compute ageftedu = -1. if (actagel2 < 16) ageftedu = 1. if (actagel2 = 16) ageftedu = 2. if (actagel2 = 17 | actagel2 = 18) ageftedu = 3. if (actagel2 = 19 | actagel2 = 20 | actagel2 = 21) ageftedu = 4. if (actagel2 > 21 and agelfte2 = 1) ageftedu = 5. if (missing(actagel2) and agelfte2 = 2) ageftedu = 6. missing values ageftedu (-1). variable labels ageftedu 'age CM left ft continuous education'. value labels ageftedu 1'before 16' 2'at 16' 3'post 16' 4'post 18' 5'post 21' 6'still in ft edu'. frequencies variables=ageftedu. age CM left full time education accounting for a 'gap' of no more than 3 years **two cases said had an additional period of education but did not say when finished, although they were not still in ft education - for these two cases I used the age of leaving ft education in ACTAGEL2 variable. compute agelfted = actagel2. if (furthed2 = 1) agelfted = lftmore2. if (agelfte2 = 2) agelfted = 99. if (serial = 109993 | serial = 217003) agelfted = actagel2. variable labels agelfted 'age CM finally left ft education'. value labels agelfted 99'still in ed'. frequencies variables=agelfted. compute ageledu = -1. if (agelfted < 16) ageledu = 1. if (agelfted = 16) ageledu = 2. if (agelfted = 17 | agelfted = 18) ageledu = 3. if (agelfted = 19 | agelfted = 20 | agelfted = 21) ageledu = 4. if (agelfted > 21) ageledu = 5. if (agelfted = 99) ageledu = 6. missing values ageledu (-1). variable labels ageledu 'age CM finally left ft education'. value labels ageledu 1'before 16' 2'at 16' 3'post 16' 4'post 18' 5'post 21' 6'still in ft edu'. frequencies variables=ageftedu ageledu. age CM partner left full-time education frequencies


variables= plefted. missing values plefted (98,99). recode plefted (14,15 = 1) (16 = 2) (17,18 = 3) (19 thru 21 = 4) (22 thru 46 = 5) (sysmis, 98, 99 = -1) into plefted1. missing values plefted1 (-1). variable labels plefted1 'age CM partner finally left education'. value labels plefted1 1'before 16' 2'at 16' 3'post 16' 4'post 18' 5'post 21'. frequencies variables= plefted1.


MATHS QUESTIONS variable system-missing user-missing: 8 (^



n % n % n % mathsprb 66 0.3 4 0 21 0.1 22589 549 (2.4%) had difficulty/were not able to work out change from £10. However, only 426 (78% of 549) asked additional 5 questions on type of maths difficulties CM had: mprbtype 426 addup 426 subtract 426 multiply 1 0 425 subtract 1 0 425 mathcors 66 0.3 4 0 20 0.1 22590 429 (1.9%) had done a course to improve their maths. 429 asked if doing course now or previously mathsnow 22251 98.1 429 mathimp 139 0.6 3 0 22538 6234 (27.7%) ever wanted to improve their maths. 6234 asked multi-coded questions i) why they wanted to improve their maths and ii) where/how they would improve them mthlike1 – mthlike5

16446 72.5 2 6232

mthplac1-mthplac9

16446 72.5 ^4 6230

datesprb 1377 6.1 1 0 15 0.1 21271 696 (3.3%) had difficulty/were not able to work out dates/use a calendar mathconf 11421 50.4 3 0 20 0.1 11236 1536 (13.7%) not confident helping their child(ren) with maths mathskid 11421 50.4 3 0 16 0.1 11240 1093 (9.7%) did not help their child(ren) with maths partmath 12733 56.1 4 0 22 0.1 9921 1439 (14.5%) of CM partner did not help their child(ren) with maths


Derived MATHS variables variable system-missing user-missing: 8 (^



n % n % n % bsamd 1394 6.1 21284 4 category variable combining information given in mathsprb and datesprb. Only for CMs who answered both questions. CM reporting ‘difficulties’ or ‘no, can’t do’ treated as one category. Values: 1’no difficulties’ 2’maths difficulties’ 3’date/calendar difficulties’ 4’maths and 3’date/calendar difficulties’. bsamp1 22254 98.1 426 6 category variable counting the specific maths problems of CMs who had reported maths difficulties in mathsprb. Included mprbtype (recognising numbers), addup, subtract, multiply, and divide. Values: 0’not these problems’ 1’one problem’ 2’two problems’ 3’three problems’ 4’four problems’ 5’all five problems’. bsamp2 22254 98.1 426 4 category variable counting the specific maths problems of CMs who had reported maths difficulties in mathsprb .Included addup, subtract, multiply, and divide. Values: 0’not these problems’ 1’one problem’ 2’two problems’ 3’three problems’ 4’all four problems’. Derived variables from multi-coded mthlike1-mthlike5: reasons why CM wants to improve their maths: for their children, to get a job, to get a promotion, get a better job, for self satisfaction. impmkid impmgjob impmprom impmbjob impmself

16448 72.5 6232

Derived variables from multi-coded mthplac1-mthplac9: means by which CM would improve their maths: on a day college course, on an evening college course, on a weekend college course, on ANY college course, at a community centre, by using local library resources, by using PC packages, by using programmes on TV, by using programmes on the radio, by using books at home. domcolld domcolle domcollw domcoll domcomc domllib dompc domtv domradio dombook

16450 72.5 6230


SPSS SYNTAX missing values mathsprb mprbtype mathcors mthimp addup subtract multiply divide datesprb mathconf mathskid mathsnow partmath (8,9). Maths and Date problems compute bsamd = -1. if (mathsprb = 1 and datesprb = 1) bsamd = 0. if (mathsprb >= 2 and datesprb = 1) bsamd = 1. if (mathsprb = 1 and datesprb >= 2) bsamd = 2. if (mathsprb >= 2 and datesprb >= 2) bsamd = 3. missing values bsamd (-1). variable labels bsamd '2000: CM has maths and calendar problems'. value labels bsamd 0'no probs' 1'maths diffs' 2'date diffs' 3'maths and date diffs'. frequencies variables=bsamd. Number of math problems. do if (mprbtype >= 1). count bsamp1 = mprbtype addup subtract multiply divide (1). end if. variable labels bsamp1 '2000: CM has probs with math calculations'. value labels bsamp1 0'no problems' 1'one problem' 2'two problems' 3'three problems' 4'four problems' 5’all five problems’. do if (mprbtype >= 1). count bsamp2 = addup subtract multiply divide (1). end if. variable labels bsamp2 '2000: CM has probs with 4 types math calculations'. value labels bsamp2 0'no problems' 1'one problem' 2'two problems' 3'three problems' 4'all four problems'. frequencies variables= bsamp1 bsamp2. Reasons why want to improve math skills missing values mthlike1 mthlike2 mthlike3 mthlike4 mthlike5 (8,9). do if (mthlike1 > 0). compute impmkid = 0. compute impmgjob = 0. compute impmprom = 0. compute impmbjob = 0. compute impmself = 0. do repeat x = mthlike1 mthlike2 mthlike3 mthlike4 mthlike5 . if (x = 1) impmkid = 1. if (x = 2) impmgjob = 1. if (x = 3) impmprom = 1. if (x = 4) impmbjob = 1. if (x = 5) impmself = 1. end repeat. end if. variable labels impmkid 'CM wants to improve maths to help their children'. variable labels impmgjob 'CM wants to improve maths to help get a job'.


variable labels impmprom 'CM wants to improve maths to help get a promotion at work'. variable labels impmbjob 'CM wants to improve maths to help get a better job'. variable labels impmself 'CM wants to improve maths to study for own satisfaction'. value labels impmkid impmgjob impmprom impmbjob impmself 0'not this reason' 1'yes, for this reason'. frequencies variables=impmkid impmgjob impmprom impmbjob impmself . Place where CM would go to improve math skills missing values mthplac1 (99). do if (mthplac1 > 0). compute domcolld = 0. compute domcolle = 0. compute domcollw = 0. compute domcoll = 0. compute domcomc = 0. compute domllib = 0. compute dompc = 0. compute domtv = 0. compute domradio = 0. compute dombook = 0. do repeat x = mthplac1 mthplac2 mthplac3 mthplac4 mthplac5 mthplac6 mthplac7 mthplac8 mthplac9. if (x = 1) domcolld = 1. if (x = 2) domcolle = 1. if (x = 3) domcollw = 1. if (x <= 3) domcoll = 1. if (x = 4) domcomc = 1. if (x = 5) domllib = 1. if (x = 6) dompc = 1. if (x = 7) domtv = 1. if (x = 8) domradio = 1. if (x = 9) dombook = 1. end repeat. end if. variable labels domcolld 'CM would improve their maths on a DAY college course'. variable labels domcolle 'CM would improve their maths on an EVENING college course'. variable labels domcollw 'CM would improve their maths on a WEEKEND college course'. variable labels domcoll 'CM would improve their maths on ANY college course'. variable labels domcomc 'CM would improve their maths at a community centre'. variable labels domllib 'CM would improve their maths by using local library resources'. variable labels dompc 'CM would improve their maths by using PC packages'. variable labels domtv 'CM would improve their maths by using programmes on TV'. variable labels domradio 'CM would improve their maths by using programmes on the RADIO'. variable labels dombook 'CM would improve their maths by using books at home'. value labels domcolld domcolle domcollw domcoll domcomc domllib dompc domtv domradio dombook 0'not using this method' 1'this method'. frequencies variables=domcolld domcolle domcollw domcoll domcomc domllib dompc domtv domradio dombook. (see COURSE section for derived reading course variable)


READING QUESTIONS variable system-missing user-missing: 8 (^



n % n % n % readprb1 1377 6.1 1 0 21302 679 (3.2%) had difficulty/were not able read/understand magazine/newspaper text. 680 asked additional 2 questions on other reading difficulties (includes CM with value of 9 in readprb1) readprb2 22000 97.0 680 readprb3 22003 97.0 3 0 677 readprb4 1377 6.1 3 0 21300 Asks if CM feels ability to read paperwork has improved over the last decade. 789 (3.7%) feel it has got worse. readcors 66 0.3 4 0 20 0.1 22590 197 (0.9%) had done a course to improve their reading. 197 asked if doing course now or previously readnow 22483 99.1 197 readimp 128 0.6 2 0 22550 1856 (8.2%) ever wanted to improve their reading. 1856 asked multi-coded questions i) why they wanted to improve their reading and ii) where/how they would improve them redlike1 – redlike5

20824 91.8 1856

redplac1-redplac9 (redplac8 and redplac9 hold no information)

20824 91.8 ^2 1854

kidrdcnf 11421 50.4 3 0 21 0.1 11235 561 (5.0%) not confident helping their child(ren) with reading readkid 11421 50.4 3 0 15 0.1 11241 822 (7.3%) did not read to their child(ren) partread 12733 56.1 4 0 21 0.1 9922 1271 (12.8%) of CM partner did not read to their child(ren)


Derived READING variables variable system-missing user-missing: 8 (^



n % n % n % bsaread1 1378 6.1 21302 4 category variable combining information given in readprb1, readprb2 and readprb3. CM reporting ‘difficulties’ or ‘no, can’t do’ treated as one category. Values: 1’no reading difficulties’ 2’I type of reading difficulty’ 3’2 types of reading difficulties’ 4’3 types of reading difficulties’. bsaread2 22003 97.0 677 3 category variable combining information given in readprb1, readprb2 and readprb3 for CMs who reported ‘difficulties’ or ‘no, can’t do’ in readprb1 and answered readprb1, readprb2 and readprb3. CM reporting ‘difficulties’ or ‘no, can’t do’ treated as one category. Values: 1’no reading difficulties’ 2’I type of reading difficulty’ 3’2 types of reading difficulties’ 4’3 types of reading difficulties’. Derived variables from multi-coded redlike1-redlike5: reasons why CM wants to improve their reading: for their children, to get a job, to get a promotion, get a better job, for self satisfaction. imprkid imprgjob imprprom imprbjob imprself

20824 91.8 1856

Derived variables from multi-coded redplac1-redplac9: means by which CM would improve their reading: on a day college course, on an evening college course, on a weekend college course, on ANY college course, at a community centre, by using local library resources, by using PC packages, by using programmes on TV, by using programmes on the radio, by using books at home. dorcolld dorcolle dorcollw dorcoll dorcomc dorllib dorpc dortv dorradio dorbook

20826 91.8 1854


SPSS SYNTAX

missing values readprb1 readprb2 readprb3 readprb4 readcors readimp partread readkid readnow kidrdcnf (8,9). Number of reading problems CM has do if (readprb1 > 0). count bsaread1 = readprb1 readprb2 readprb3 (2,3) / end if. do if (readprb1 > 1). count bsaread2 = readprb1 readprb2 readprb3 (2,3) / end if. variable labels bsaread1 'CM not able to do or has difficulties with reading tasks'. variable labels bsaread2 'CM not able to do or has difficulties with reading tasks'. frequencies variables=bsaread1 bsaread2 . Reasons CM wants to improve reading skills missing values redlike1 redlike2 redlike3 redlike4 redlike5 (8,9). do if (redlike1 > 0). compute imprkid = 0. compute imprgjob = 0. compute imprprom = 0. compute imprbjob = 0. compute imprself = 0. do repeat x = redlike1 redlike2 redlike3 redlike4 redlike5 . if (x = 1) imprkid = 1. if (x = 2) imprgjob = 1. if (x = 3) imprprom = 1. if (x = 4) imprbjob = 1. if (x = 5) imprself = 1. end repeat. end if. variable labels imprkid 'CM wants to improve reading to help their children'. variable labels imprgjob 'CM wants to improve reading to help get a job'. variable labels imprprom 'CM wants to improve reading to help get a promotion at work'. variable labels imprbjob 'CM wants to improve reading to help get a better job'. variable labels imprself 'CM wants to improve reading to study for own satisfaction'. value labels imprkid imprgjob imprprom imprbjob imprself 0'not this reason' 1'yes, for this reason'. frequencies variables=imprkid imprgjob imprprom imprbjob imprself .


Place where CM would go to improve reading skills missing values redplac1 (99). do if (redplac1 > 0). compute dorcolld = 0. compute dorcolle = 0. compute dorcollw = 0. compute dorcoll = 0. compute dorcomc = 0. compute dorllib = 0. compute dorpc = 0. compute dortv = 0. compute dorradio = 0. compute dorbook = 0. do repeat x = redplac1 redplac2 redplac3 redplac4 redplac5 redplac6 redplac7 redplac8 redplac9 . if (x = 1) dorcolld = 1. if (x = 2) dorcolle = 1. if (x = 3) dorcollw = 1. if (x <= 3) dorcoll = 1. if (x = 4) dorcomc = 1. if (x = 5) dorllib = 1. if (x = 6) dorpc = 1. if (x = 7) dortv = 1. if (x = 8) dorradio = 1. if (x = 9) dorbook = 1. end repeat. end if. variable labels dorcolld 'CM would improve their reading on a DAY college course'. variable labels dorcolle 'CM would improve their reading on an EVENING college course'. variable labels dorcollw 'CM would improve their reading on a WEEKEND college course'. variable labels dorcoll 'CM would improve their reading on ANY college course'. variable labels dorcomc 'CM would improve their reading at a community centre'. variable labels dorllib 'CM would improve their reading by using local library resources'. variable labels dorpc 'CM would improve their reading by using PC packages'. variable labels dortv 'CM would improve their reading by using programmes on TV'. variable labels dorradio 'CM would improve their reading by using programmes on the RADIO'. variable labels dorbook 'CM would improve their reading by using books at home'. value labels dorcolld dorcolle dorcollw dorcoll dorcomc dorllib dorpc dortv dorradio dorbook 0'not using this method' 1'this method'. frequencies variables=dorcolld dorcolle dorcollw dorcoll dorcomc dorllib dorpc dortv dorradio dorbook. (see COURSE section for derived reading course variable)


WRITING QUESTIONS variable system-missing user-missing: 8 (^



n % n % n % writeprb 1377 6.1 1 0 21302 1087 (3.2%) had difficulty/were not able to write a thank-you letter to a friend. 1087 asked additional question on spelling difficulties (includes CM with value of 9 in writeprb) wprbtype 21593 95.2 1 0.1 1086 70 (6.4%) never tried to write a thank-you letter to a friend. 1017 asked additional questions on handwriting legibility and difficulties putting down in words what want to say hwritprb 21663 95.5 1 0.1 1016 wordsprb 21663 95.5 1017 writcors 66 0.3 4 0 21 0.1 22589 315 (1.4%) had done a course to improve their writing. 315 asked if doing course now or previously writenow 22365 98.6 315 writimp 138 0.6 3 0 22539 2636 (11.7%) ever wanted to improve their writing. 2636 asked multi-coded questions i) why they wanted to improve their reading and ii) where/how they would improve them wrilike1 – wrilike5

20044 88.4 1 0 2635

wriplac1-wriplac9

20044 88.4 ^5 2631

writconf 11421 50.4 4 0 26 0.1 11229 806 (7.2%) not confident helping their child(ren) with writing writekid 11421 50.4 3 0 14 0.1 11242 1440 (12.8%) did not help their child(ren) with writing partwrit 12733 56.1 4 0 21 0.1 9922 1687 (17.0%) of CM partner did help their child(ren) with writing


Derived WRITING variables variable system-missing user-missing: 8 (^



n % n % n % bsawrit1 1379 6.1 21301 8 category variable combining information given in writeprb, hwritprb, wprbtype and wordsprb. CM reporting ‘difficulties’ or ‘no, can’t do’ treated as one category. Values: 0'no probs' 1'writing spelling handwriting articulating' 2'writing spelling handwriting' 3'writing spelling articulating' 4'writing handwriting articulating' 5'writing spelling' 6'writing articulating' 7'writing handwriting' 8'writing only'. bsawrit2 21594 95.2 1086 4 category variable counting the specific writing problems of CMs who had reported ‘difficulties’ or ‘no, can’t do’ in writeprb. Included writeprb, wprbtype (spelling), hwritprb (handwriting), and wordsprb (articulating). CM reporting ‘difficulties’ or ‘no, can’t do ’in writeprb treated as one category. Values: 1’1 difficulty’ 2’2 types of writing difficulties’ 3’3 types of writing difficulties’ 4’4 types of writing difficulties’. Derived variables from multi-coded redlike1-redlike5: reasons why CM wants to improve their reading: for their children, to get a job, to get a promotion, get a better job, for self satisfaction. Values: 0'not this reason' 1'yes, for this reason'. impwkid impwgjob impwprom impwbjob impwself

20045 88.4 2635

Derived variables from multi-coded redplac1-redplac9: means by which CM would improve their reading: on a day college course, on an evening college course, on a weekend college course, on ANY college course, at a community centre, by using local library resources, by using PC packages, by using programmes on TV, by using programmes on the radio, by using books at home. Values: 0'not using this method' 1'this method'. dowcolld dowcolle dowcollw dowcoll dowcomc dowllib dowpc dowtv dowradio dowbook

20049 88.4 2631


SPSS SYNTAX Type of writing difficulties Cm has missing values writeprb hwritprb wprbtype wordsprb writimp writcors writconf writekid writenow partwrit (8,9). compute bsawrit1 = -1. if (writeprb = 1) bsawrit1 = 0. if (writeprb >= 2 and (wprbtype = 1 | wprbtype = 3) and (missing(hwritprb) | hwritprb = 1) and (missing(wordsprb) | wordsprb = 1)) bsawrit1 = 1. if (writeprb >= 2 and (wprbtype = 1 | wprbtype = 3) and (missing(hwritprb) | hwritprb = 1) and wordsprb = 2) bsawrit1 = 2. if (writeprb >= 2 and (wprbtype = 1 | wprbtype = 3) and hwritprb = 2 and (missing(wordsprb) | wordsprb = 1))bsawrit1 = 3. if (writeprb >= 2 and wprbtype = 2 and (missing(hwritprb) | hwritprb = 1) and (missing(wordsprb) | wordsprb = 1)) bsawrit1 = 4. if (writeprb >= 2 and (wprbtype = 1 | wprbtype = 3) and hwritprb = 2 and wordsprb = 2) bsawrit1 = 5. if (writeprb >= 2 and wprbtype = 2 and hwritprb = 2 and (missing(wordsprb) | wordsprb = 1)) bsawrit1 = 6. if (writeprb >= 2 and wprbtype = 2 and (missing(hwritprb) | hwritprb = 1) and wordsprb = 2) bsawrit1 = 7. if (writeprb >= 2 and wprbtype = 2 and hwritprb = 2 and wordsprb = 2) bsawrit1 = 8. missing values bsawrit1 (-1). variable labels bsawrit1 'type of writing problems CM has'. value labels bsawrit1 0'no probs' 1'writing spelling handwriting articulating' 2'writing spelling handwriting' 3'writing spelling articulating' 4'writing handwriting articulating' 5'writing spelling' 6'writing articulating' 7'writing handwriting' 8'writing only'. Number of writing difficulties CM has recode writeprb (1=0) (2,3=1) into write. recode wprbtype (2=0) (1,3=1) into wprb. do if (write = 1). count bsawrit2 = write wprb hwritprb wordsprb (1). end if. variable labels bsawrit2 'Number of writing difficulties CM has'. value labels bsawrit2 0'no probs' 1'one difficulty' 2'two difficulties' 3'three difficulties' 4'all four difficulties'. frequencies variables=bsawrit1 bsawrit2.


Reasons why CM wants to improve writing skills frequencies variables=writimp wrilike1 wrilike2 wrilike3 wrilike4 wrilike5 wriplac1 wriplac2 wriplac3 wriplac4 wriplac5 wriplac6 wriplac7 wriplac8 wriplac9 . missing values wrilike1 wrilike2 wrilike3 wrilike4 wrilike5 (8,9). do if (wrilike1 > 0). compute impwkid = 0. compute impwgjob = 0. compute impwprom = 0. compute impwbjob = 0. compute impwself = 0. do repeat x = wrilike1 wrilike2 wrilike3 wrilike4 wrilike5 . if (x = 1) impwkid = 1. if (x = 2) impwgjob = 1. if (x = 3) impwprom = 1. if (x = 4) impwbjob = 1. if (x = 5) impwself = 1. end repeat. end if. variable labels impwkid 'CM wants to improve writing to help their children'. variable labels impwgjob 'CM wants to improve writing to help get a job'. variable labels impwprom 'CM wants to improve writing to help get a promotion at work'. variable labels impwbjob 'CM wants to improve writing to help get a better job'. variable labels impwself 'CM wants to improve writing to study for own satisfaction'. value labels impwkid impwgjob impwprom impwbjob impwself 0'not this reason' 1'yes, for this reason'. frequencies variables=impwkid impwgjob impwprom impwbjob impwself . Place where CM would go to improve writing skills missing values wriplac1 (99). do if (wriplac1 > 0). compute dowcolld = 0. compute dowcolle = 0. compute dowcollw = 0. compute dowcoll = 0. compute dowcomc = 0. compute dowllib = 0. compute dowpc = 0. compute dowtv = 0. compute dowradio = 0. compute dowbook = 0. do repeat x = wriplac1 wriplac2 wriplac3 wriplac4 wriplac5 wriplac6 wriplac7 wriplac8 wriplac9. if (x = 1) dowcolld = 1. if (x = 2) dowcolle = 1. if (x = 3) dowcollw = 1. if (x <= 3) dowcoll = 1. if (x = 4) dowcomc = 1. if (x = 5) dowllib = 1. if (x = 6) dowpc = 1. if (x = 7) dowtv = 1. if (x = 8) dowradio = 1.


if (x = 9) dowbook = 1. end repeat. end if. variable labels dowcolld 'CM would improve their writing on a DAY college course'. variable labels dowcolle 'CM would improve their writing on an EVENING college course'. variable labels dowcollw 'CM would improve their writing on a WEEKEND college course'. variable labels dowcoll 'CM would improve their writing on ANY college course'. variable labels dowcomc 'CM would improve their writing at a community centre'. variable labels dowllib 'CM would improve their writing by using local library resources'. variable labels dowpc 'CM would improve their writing by using PC packages'. variable labels dowtv 'CM would improve their writing by using programmes on TV'. variable labels dowradio 'CM would improve their writing by using programmes on the RADIO'. variable labels dowbook 'CM would improve their writing by using books at home'. value labels dowcolld dowcolle dowcollw dowcoll dowcomc dowllib dowpc dowtv dowradio dowbook 0'not using this method' 1'this method'. frequencies variables=dowcolld dowcolle dowcollw dowcoll dowcomc dowllib dowpc dowtv dowradio dowbook writenow. (see COURSE section for derived writing course variable)


DIFFICULTIES OF HAVING BASIC SKILLS PROBLEMS variable system-missing user-missing: 8 (^



n % n % n % getjob 21047 92.8 21 0.1 1612 620 (38.5%) CMs with 3R problem felt affected them getting new job copejob 21047 92.8 15 0.1 1618 391 (24.2%) CMs with 3R problem felt affected them coping in current job gtpromot 21047 92.8 38 0.2 1595 570 (35.7%) CMs with 3R problem felt affected them getting a promotion copehmbs 21047 92.8 6 0 1627 253 (15.6%) CMs with 3R problem felt affected them in managing household business helpkids 21047 92.8 3 0 1630 419 (25.7%) CMs with 3R problem felt affected them being able to help their children lear. In addition, 409 (25.1%) never helped their children to read or learn copeleis 21047 92.8 4 0 1629 341 (20.9%) CMs with 3R problem felt affected them pursuing other interests


SKILLS (self-completed) variable system-missing user-missing: 8 (^



n % n % n % SKILL1A 283 1.2 3 5 22344 CM reported ‘good’ ‘okay’ or ‘poor’ communicating skills. 21544 (96%) reported if they used this skill at work. SKILL1B 1136 5.0 1 21543 SKILL2A 283 1.2 3 4 22268 CM reported ‘good’ ‘okay’ or ‘poor’ number/calculation skills. 21477 (96%) reported if they used this skill at work. SKILL2B 2103 5.3 1 21476 SKILL3A 283 1.2 3 7 19726 CM reported ‘good’ ‘okay’ or ‘poor’ computer/IT skills. 19173 (86%) reported if they used this skill at work. SKILL3B 3507 15.5 1 19172 SKILL4A 283 1.2 3 7 22219 CM reported ‘good’ ‘okay’ or ‘poor’ team working skills. 21469 (96%) reported if they used this skill at work. SKILL4B 1211 5.3 3 21466 SKILL5A 283 1.2 3 6 22297 CM reported if they were ‘good’ ‘okay’ or ‘poor’ at learning new skills. 21517 (96%) reported if they used this skill at work. SKILL5B 1163 5.1 1 21516 SKILL6A 283 1.2 3 7 22308 CM reported ‘good’ ‘okay’ or ‘poor’ problem solving skills. 21515 (96%) reported if they used this skill at work. SKILL6B 1165 5.1 1 21514 SKILL7A 283 1.2 3 6 21909 CM reported ‘good’ ‘okay’ or ‘poor’ tool use skills. 21149 (94%) reported if they used this skill at work. SKILL7B 1531 6.8 1 21148 SKILL8A 283 1.2 3 19 0.1% 20975 CM reported ‘good’ ‘okay’ or ‘poor’ caring skills. 20217 (90%) reported if they used this skill at work. SKILL8B 2463 10.9 1 21216 SKILL9A 283 1.2 3 5 21228 CM reported ‘good’ ‘okay’ or ‘poor’ finance/accounting skills. 20523 (92%) reported if they used this skill at work. SKILL9B 2157 9.5 1 20522


COURSES variable system-missing user-missing: 8 (^



n % n % n % failqual 66 0.3 3 34 0.2 22577 3404 (8.8%) failed/cot completed course. 3404 asked how many courses they had failed/not completed numfqual 19276 84.9 ^3 3401 yts (bcs70 only) 11454 50.5 3 12 0.1 11211 3303 (29.5%) CMs had done a yts course. 3303 asked how many they had done, and if they were on a course now numyts 19377 85.4 ^1 ^1 3301 ytsnow 19377 85.4 1 3302 othgov 66 0.3 4 22 0.1 22588 550 (2.4%) CMs had done an ‘other’ type of Government course. 550 asked how many they had done, and if they were on a course now numgov 22130 97.6 ^3 0.5 547 govnow 22130 97.6 550 aptrain 66 0.3 4 24 0.1 22586 329 (1.5 %) CMs had done a modern apprenticeship course. 329 asked how many they had done, and if they were on a course now numap 22351 98.5 329 apnow 22351 98.5 329 actrain 66 0.3 4 23 0.1 22587 352 (1.6%) CMs had done an access course. 3303 asked how many they had done, and if they were on a course now numac 22328 98.4 352 acnow 22328 98.4 352 wrktrain 66 0.3 4 24 0.1 22586 7871 (34.7%) CMs had done a work-related training course. 7871 asked how many they had done, and if they were on a course now numwrktr 14809 65.3 ^66 0.8 7805 wrktrnow 14809 65.3 1 0 7870 leiscors 66 0.3 5 23 0.1 22586 5528 (24.4%) CMs had done a course for interest or leisure. 5528 asked how many they had done, and if they were on a course now numleis 17152 75.6 ^4 0.1 5524 leisnow 17152 75.6 5528


Derived COURSE variables variable system-missing user-missing: 8 (^



n % n % n % yts1 11470 50.6 11210 3 category variable combining information from yts and ytsnow. Values: 0'no' 1'yes, previously' 2'yes, now'. gov1 94 0.4 22588 3 category variable combining information from othgov and govnow. Values: 0'no' 1'yes, previously' 2'yes, now'. apt1 92 0.4 22586 3 category variable combining information from aptrain and apnow. Values: 0'no' 1'yes, previously' 2'yes, now'. access1 95 0.4 22587 3 category variable combining information from actrain and acnow. Values: 0'no' 1'yes, previously' 2'yes, now'. wrt1 93 0.4 22585 3 category variable combining information from wrktrain and wrktrnow. Values: 0'no' 1'yes, previously' 2'yes, now'. leis1 94 0.4 22586 3 category variable combining information from leiscors and leisnow. Values: 0'no' 1'yes, previously' 2'yes, now'. read1 90 0.4 22590 3 category variable combining information from readcors and readnow. Values: 0'no' 1'yes, previously' 2'yes, now'. writ1 91 0.4 22589 3 category variable combining information from writcors and writenow. Values: 0'no' 1'yes, previously' 2'yes, now'. math1 90 0.4 22590 3 category variable combining information from mathcors and mathsnow. Values: 0'no' 1'yes, previously' 2'yes, now'.


SPSS SYNTAX missing values failqual yts ytsnow othgov govnow aptrain apnow actrain acnow wrktrain wrktrnow leiscors leisnow readcors readnow writcors writenow mathcors mathsnow (8,9). missing values numfqual numyts numgov numap numac numwrktr numleis numread numwrite nummaths (98,99). compute yts1 = -1. if (yts = 2) yts1 = 0. if (ytsnow = 2) yts1 = 1. if (ytsnow = 1) yts1 = 2. missing values yts1 (-1). variable labels yts1 'CM been on YTS?'. value labels yts1 0'no' 1'yes, previously' 2'yes, now'. compute gov1 = -1. if (othgov = 2) gov1 = 0. if (govnow = 2) gov1 = 1. if (govnow = 1) gov1 = 2. missing values gov1 (-1). variable labels gov1 'CM been on other government course?'. value labels gov1 0'no' 1'yes, previously' 2'yes, now'. compute apt1 = -1. if (aptrain = 2) apt1 = 0. if (apnow = 2) apt1 = 1. if (apnow = 1) apt1 = 2. missing values apt1 (-1). variable labels apt1 'CM been on modern apprenticeship?'. value labels apt1 0'no' 1'yes, previously' 2'yes, now'. compute access1 = -1. if (actrain = 2) access1 = 0. if (acnow = 2) access1 = 1. if (acnow = 1) access1 = 2. missing values access1 (-1). variable labels access1 'CM been on access course'. value labels access1 0'no' 1'yes, previously' 2'yes, now'. compute wrt1 = -1. if (wrktrain = 2) wrt1 = 0. if (wrktrnow = 2) wrt1 = 1. if (wrktrnow = 1) wrt1 = 2. missing values wrt1 (-1). variable labels wrt1 'CM been on work-related training course'. value labels wrt1 0'no' 1'yes, previously' 2'yes, now'. compute leis1 = -1. if (leiscors = 2) leis1 = 0. if (leisnow = 2) leis1 = 1. if (leisnow = 1) leis1 = 2. missing values leis1 (-1). variable labels leis1 'CM been on leisure course'. value labels leis1 0'no' 1'yes, previously' 2'yes, now'.


compute readc1 = -1. if (readcors = 2) readc1 = 0. if (readnow = 2) readc1 = 1. if (readnow = 1) readc1 = 2. missing values readc1 (-1). variable label readc1 'CM been on reading improvement course'. value labels readc1 0'no' 1'yes, previously' 2'yes, now'. compute writ1 = -1. if (writcors = 2) writ1 = 0. if (writenow = 2) writ1 = 1. if (writenow = 1) writ1 = 2. missing values writ1 (-1). variable labels writ1 'CM been on writing improvement course'. value labels writ1 0'no' 1'yes, previously' 2'yes, now'. compute math1 = -1. if (mathcors = 2) math1 = 0. if (mathsnow = 2) math1 = 1. if (mathsnow = 1) math1 = 2. missing values math1 (-1). variable labels math1 'CM been on basic maths improvement course'. value labels math1 0'no' 1'yes, previously' 2'yes, now'. frequencies variables=yts1 apt1 gov1 wrt1 access1 leis1 read1 writ1 math1.


COMPUTER USE AT HOME and WORK variable system-missing user-missing: 8 (^



n % n % n % pchome 66 0.3 4 20 0.1 22590 13379 (59.0%) CM had a PC at home. 13379 asked how often they used the hpcuse 9301 41.0 4 0 13375 11289 (84.4%) CMs used the PC at home. 11289 asked about the ways they use a PC howuse01-howuse33 (howuse18-howuse33 hold no information)

11391 50.2 ^8 0.1 11281

pcwork 3915 17.3 2 9 0 18754 12026 (64.1%) CM use a PC at work. 12026 asked how often use a PC and the ways they use a PC wpcuse 10654 47.0 1 12025 howuse34-howuse46

10654 47.0 ^2 0 12024

Derived HOME COMPUTER variables



valid n

n % n % n % pchome1 94 0.4 22586 6 category variable combining information from pchome and hpcuse. Values: 0'no' 1'yes, but do not use it' 2'yes, use it less than once a week' 3'yes, use it once a week' 4'yes, use it 2-4 times a week' 5'yes, use it daily' Derived variables from multi-coded howuse01-howuse33. For 11289 (50.0%) CMs who have and use a home computer. Ways CM uses home computer: for word processing, WWW, e-mail, data analysis, data base work, design packages, playing games, sending/receiving faxes, using CD ROM or Encyclopedia, composing music, listening to music, photography, programming, managing home finances, spreadsheets, webb design, scanning, other things, unspecified things. Values: 0'not use home PC this way' 1'yes, uses home PC this way'. hwp hwww hemail hdatan hdatab hdesign hgames hfax hencyrom hmusicc hmusicl hphoto hprog hhomefin hspread hwebdes hscan hother hunspec

11391 50.2 11289

Derived WORK COMPUTER variables variable system-missing user-missing: 8 (^



n % n % n % pcwork1 3927 17.3 18753 5 category variable combining information from pcwork and wpcuse. Values 0'no' 1'yes, use it less than once a week' 2'yes, use it once a week' 3'yes, use it 2-4 times a week' 4'yes, use it daily'.


Derived variables from multi-coded howuse34-howuse46. For 12025 (64.1%) CMs who use a PC at work. Ways CM uses work computer: for word processing, WWW, e-mail, data analysis, data base work, design packages, playing games, sending/receiving faxes, using CD ROM or Encyclopedia, composing music, listening to music, photography, programming, spreadsheets, webb design, scanning, other things, unspecified things. Values: 0'not use work PC this way' 1'yes, uses work PC this way'. wwp wwww wemail wdatan wdatab wdesign wgames wfax wencyrom wmusicc wmusicl wphoto wprog wother

10655 47.0 12025


SPSS SYNTAX Computer use at home compute pchome1 = -1. if (pchome = 2) pchome1 = 0. if (pchome = 1 and hpcuse = 1) pchome1 = 1. if (pchome = 1 and hpcuse = 5) pchome1 = 2. if (pchome = 1 and hpcuse = 4) pchome1 = 3. if (pchome = 1 and hpcuse = 3) pchome1 = 4. if (pchome = 1 and hpcuse = 2) pchome1 = 5. missing values pchome1 (-1). variable labels pchome1 'Does CM have and use a PC at home?'. value labels pchome1 0'no' 1'yes, but do not use it' 2'yes, use it less than once a week' 3'yes, use it once a week' 4'yes, use it 2-4 times a week' 5'yes, use it daily'. frequencies variables=pchome1. do if (pchome1 >= 0). compute hwp = 0. compute hwww = 0. compute hemail = 0. compute hdatan = 0. compute hdatab = 0. compute hdesign = 0. compute hgames = 0. compute hfax = 0. compute hencyrom = 0. compute hmusicc = 0. compute hmusicl = 0. compute hphoto = 0. compute hprog = 0. compute hhomefin = 0. compute hspread = 0. compute hwebdes = 0. compute hscan = 0. compute hother = 0. compute hunspec = 0. do repeat x = howuse01 howuse02 howuse03 howuse04 howuse05 howuse06 howuse07 howuse08 howuse09 howuse10 howuse11 howuse12 howuse13 howuse14 howuse15 howuse16 howuse17 . if (x = 1) hwp = 1. if (x = 2) hwww = 1. if (x = 3) hemail = 1. if (x = 4) hdatan = 1. if (x = 5) hdatab = 1. if (x = 6) hdesign = 1. if (x = 7) hgames = 1. if (x = 8) hfax = 1. if (x = 9) hencyrom = 1. if (x = 10) hmusicc = 1. if (x = 11) hmusicl = 1. if (x = 12) hphoto = 1. if (x = 13) hprog = 1. if (x = 14) hhomefin = 1. if (x = 15) hspread = 1. if (x = 16) hwebdes = 1. if (x = 17) hscan = 1. if (x = 18) hother = 1. if (x = 19) hunspec = 1.


end repeat. end if. variable labels hwp 'CM uses home PC for word processing'. variable labels hwww 'CM uses home PC for WWW'. variable labels hemail 'CM uses home PC for e-mail'. variable labels hdatan 'CM uses home PC for data analysis'. variable labels hdatab 'CM uses home PC for data base work'. variable labels hdesign 'CM uses home PC for design packages'. variable labels hgames 'CM uses home PC for playing games'. variable labels hfax 'CM uses home PC for sending/receiving faxes'. variable labels hencyrom 'CM uses home PC for using CD ROM or Encyclopedia'. variable labels hmusicc 'CM uses home PC for composing music'. variable labels hmusicl 'CM uses home PC for listening to music'. variable labels hphoto 'CM uses home PC for photography'. variable labels hprog 'CM uses home PC for programming'. variable labels hhomefin 'CM uses home PC for managing home finances'. variable labels hspread 'CM uses home PC for using spreadsheets'. variable labels hwebdes 'CM uses home PC for webb design'. variable labels hscan 'CM uses home PC for scanning'. variable labels hother 'CM uses home PC for other things'. variable labels hunspec 'CM uses home PC for unspecified things'. value labels hwp hwww hemail hdatan hdatab hdesign hgames hfax hencyrom hmusicc hmusicl hphoto hprog hhomefin hspread hwebdes hscan hother hunspec 0'not use home PC this way' 1'yes, uses home PC this way'. frequencies variables=pchome1 hwp hwww hemail hdatan hdatab hdesign hgames hfax hencyrom hmusicc hmusicl hphoto hprog hhomefin hspread hwebdes hscan hother hunspec . Computer use at work compute pcwork1 = -1. if (pcwork = 2) pcwork1 = 0. if (pcwork = 1 and wpcuse = 4) pcwork1 = 1. if (pcwork = 1 and wpcuse = 3) pcwork1 = 2. if (pcwork = 1 and wpcuse = 2) pcwork1 = 3. if (pcwork = 1 and wpcuse = 1) pcwork1 = 4. missing values pcwork1 (-1). variable labels pcwork1 'Does CM use a PC at work?'. value labels pcwork1 0'no' 1'yes, use it less than once a week' 2'yes, use it once a week' 3'yes, use it 2-4 times a week' 4'yes, use it daily'. frequencies variables=pcwork1. do if (pcwork1 >= 0). compute wwp = 0. compute wwww = 0. compute wemail = 0. compute wdatan = 0. compute wdatab = 0. compute wdesign = 0. compute wgames = 0. compute wfax = 0. compute wencyrom = 0. compute wmusicc = 0.


compute wmusicl = 0. compute wphoto = 0. compute wprog = 0. compute wother = 0. do repeat x = howuse34 howuse35 howuse36 howuse37 howuse38 howuse39 howuse40 howuse41 howuse42 howuse43 howuse44 howuse45 howuse46 . if (x = 1) wwp = 1. if (x = 2) wwww = 1. if (x = 3) wemail = 1. if (x = 4) wdatan = 1. if (x = 5) wdatab = 1. if (x = 6) wdesign = 1. if (x = 7) wgames = 1. if (x = 8) wfax = 1. if (x = 9) wencyrom = 1. if (x = 10) wmusicc = 1. if (x = 11) wmusicl = 1. if (x = 12) wphoto = 1. if (x = 13) wprog = 1. if (x = 14) wother = 1. end repeat. end if. variable labels wwp 'CM uses work PC for word processing'. variable labels wwww 'CM uses work PC for WWW'. variable labels wemail 'CM uses work PC for e-mail'. variable labels wdatan 'CM uses work PC for data analysis'. variable labels wdatab 'CM uses work PC for data base work'. variable labels wdesign 'CM uses work PC for design packages'. variable labels wgames 'CM uses work PC for playing games'. variable labels wfax 'CM uses work PC for sending/receiving faxes'. variable labels wencyrom 'CM uses work PC for using CD ROM or Encyclopedia'. variable labels wmusicc 'CM uses work PC for composing music'. variable labels wmusicl 'CM uses work PC for listening to music'. variable labels wphoto 'CM uses work PC for photography'. variable labels wprog 'CM uses work PC for programming'. variable labels wother 'CM uses work PC for other things'. value labels wwp wwww wemail wdatan wdatab wdesign wgames wfax wencyrom wmusicc wmusicl wphoto wprog wother 0'not use' 1'uses'. frequencies variables=wwp wwww wemail wdatan wdatab wdesign wgames wfax wencyrom wmusicc wmusicl wphoto wprog wother .


DRUG QUESTIONS variable system-missing user-missing: 8 (^



n % n % n % CANNABIS 283 1.2 13 0.1 6 0.0 22378 ECSTASY 283 1.2 12 0.1 7 0.0 22378 AMPHET 283 1.2 11 0 6 0.0 22380 LSD 283 1.2 11 0 6 0.0 22380 POPPER 283 1.2 11 0 8 0.0 22378 MAGMUSH 283 1.2 11 0 7 0.0 22379 COCAINE 283 1.2 11 0 7 0.0 22379 TEMAZ 283 1.2 11 0 9 0.0 22377 SEMERON 283 1.2 11 0 10 0.0 22376 KETAMINE 283 1.2 11 0 8 0.0 22378 CRACK 283 1.2 11 0 6 0.0 22380 HEROIN 283 1.2 11 0 6 0.0 22380 METHAD 283 1.2 11 0 6 0.0 22380 OTHDRUG 283 1.2 11 0 7 0.0 22379 DRUG-DRUG6: multi-coded string variables for the 614 CMs who answered ‘yes’ to othdrug. 22378 are ‘numeric missing’ leaving just 302 individual answers in drug, 37 in drug2, 13 in drug3, 8 in drug4, 4 in drug5, 2 in drug6. Each answer needs to be converted from string to numeric.

DERIVED DRUG VARIABLES variable system-missing user-missing: 8 (^



n % n % n % If CMs said they had taken SEMERON they were omitted from derived variables on illegal drug use. 43 (0.2%) said they had taken the drug SEMERON. anydrug1 347 1.5 22333 Takes information from the 12 individual illegal drug questions (information in othdrug not coded yet). Counts the number of illegal drugs CM has taken in the last 12 months. 13.5% of CMs have taken at least 1 illegal drug in last 12 months. drugs1 347 1.5 22333 Collapses anydrug1 variable into 0’not taken any drug’ 1’yes, taken at least 1 illegal drug in last 12 months’. anydrug2 347 1.5 22333 Takes information from the 12 individual illegal drug questions (information in othdrug not coded yet). Counts the number of illegal drugs CM has ever taken. 43.5% of CMs have taken at least 1 illegal drug in their lifetime. drugs2 347 1.5 22333 Collapses anydrug2 variable into 0’never taken an illegal drug’ 1’yes, taken at least 1 illegal drug in lifetime’.


SPSS SYNTAX missing values cannabis ecsacy amphet lsd popper magmush cocaine temaz semeron ketamine crack heroin methad othdrug (8,9). do if (semeron = 1). count anydrug1 = cannabis ecsacy amphet lsd popper magmush cocaine temaz ketamine crack heroin methad (3) / anydrug2 = cannabis ecsacy amphet lsd popper magmush cocaine temaz ketamine crack heroin methad (2, 3). end if. variable labels anydrug1 ' CM taken any illegal drug in last 12 months'. variable labels anydrug2 'CM ever taken any illegal drug'. frequencies variables=anydrug1 anydrug2. recode anydrug1 (0=0) (1 thru 12=1) into drugs1 . recode anydrug2 (0=0) (1 thru 12=1) into drugs2. variable labels drugs1 'CM taken any illegal drug in last 12 months'. variable labels drugs2 'CM ever taken any illegal drug'. value labels drugs1 drugs2 0'no' 1'yes'. frequencies variables=drugs1 drugs2.


DIET FOOD EXERCISE variable system-missing user-missing: 8 (^



n % n % n % fruit 66 0.3 9 28 0.1 22577 eggs 66 0.3 10 29 0.1 22575 salads 66 0.3 10 28 0.1 22576 cookdveg 66 0.3 10 26 0.1 22578 oilfried 66 0.3 11 37 0.2 22566 chops 66 0.3 11 45 0.2 22558 sweets 66 0.3 11 27 0.1 22576 cakes 66 0.3 11 26 0.1 22577 whlbread 66 0.3 11 27 0.1 22576 othbread 66 0.3 11 26 0.1 22577 redmeat 66 0.3 11 26 0.1 22577 poultry 66 0.3 11 26 0.1 22577 fish 66 0.3 11 26 0.1 22577 pulses 66 0.3 11 27 0.1 22576 veggy 66 0.3 9 26 0.1 22579 808 (3.6%) CMs vegetarian. 808 asked the type of vegetarian they were vegtype 808 spshdiet 66 0.3 9 26 0.1 22579 1061 (4.7%) CMs followed a special diet. 1061 asked the type of special diet they followed. Only 1009 (95% of 1061) asked if the diet had been recommended by a doctor diettype 1061

dietdoc 1009 exercise 66 0.3 10 27 0.1 22577 17263 (76.3%) CMs do regular exercise. 17263 asked how often they take regular exercise and if they become breathless/sweaty during exercis breathle 2 0 17261

sweat 3 0 17260


EATING PROBLEMS variable system-missing user-missing: 8 (^



n % n % n % eatprob 66 0.3 6 21 0.1 22587 729 (3.2%) CMs reported an eating disorder. 729 asked name of eating problem(s) they have had in multi-coded variables eating1- eating4 (eating4 holds no information)

21951 96.8 1 728

el1age 21951 96.8 ^5 0.7 724 el112m 21951 96.8 1 0.1 1 0.1 727 349 (47.9%) CMs had an eating disorder in last 12 months. 349 asked if they had seen a doctor re: eating disorder el1doc 22331 349 Derived EATING PROBLEMS variables



valid n

n % n % n % eatprnow 95 0.4 22585 3 category variable combining information given in eatprob and el112m. Values: 0’no eating problems’ 1’previous eating problem’ 2’current eating problem’ bulemia 93 0.4 22587 anorexia 93 22587 swallow 93 22587 otheatpr 93 22587 2 category variables. If CM gave valid answer to eatprob, used information in multi-coded eating1-eating4 variables. Values: 0’not this eating problem’ 1’yes, this eating problem


SPSS SYNTAX missing values eatprob eating1 eating2 eating3 eating4 el112m el1doc (8,9). missing values el1age (98,99). compute eatprnow = -1. if (eatprob = 2) eatprnow = 0. if (eatprob = 1 and el112m = 2) eatprnow = 1. if (eatprob = 1 and el112m = 1) eatprnow = 2. missing values eatprnow (-1). variable labels eatprnow 'CM suffers from eating disorders?'. value labels eatprnow 0’no eating problems’ 1’previous eating problem’ 2’current eating problem’. frequencies variables=eatprnow. do if (eatprob <= 2). compute bulemia = 0. compute anorexia = 0. compute swallow = 0. compute otheatpr = 0. do repeat x = eating1 eating2 eating3 eating4 . if (x = 1) bulemia = 1. if (x = 2) anorexia = 1. if (x = 3) swallow = 1. if (x = 4) otheatpr = 1. end repeat. end if. variable labels bulemia 'cm has suffered from bulemia'. variable labels anorexia 'cm has suffered from anorexia'. variable labels swallow 'cm has suffered from swallowing difficulties'. variable labels otheatpr 'cm has suffered from otheatpr'. value labels bulemia anorexia swallow otheatpr 0’not this eating problem’ 1’yes, this eating problem’. frequencies variables=bulemia anorexia swallow otheatpr .


SMOKING QUESTIONS variable system-missing user-missing: 8 (^

indicates 98; ^îndicates 998)


^îndicates 999)

valid n

n % n % n % smoking 66 0.3 8 26 0.1 22580 6222 (27.4%) CMs smoke every day. 6222 sked how many they smoke a day nofcigs 16458 72.6 ^^16 0.3 6206 6347 (28%) CMs ex-smokers or occasional smokers. 6347 asked if ever smoked cigarettes regularly exsmoker 16333 72.0 6347 4532 (71.4%) CMs had smoked regularly. 4532 asked age last smoked regularly agequit 18148 80.0 ^6 0.1 4526 othsmoke 2382 10.5 8 29 0.1 20261 5864 (25.9%) live with someone who smokes in CM home. 5864 asked whether this was spouse/partner or someone else whosmoke 16816 74.1 5864 4521 (77.1%) CMs said partner smoked. 4521 asked how many spouse/partner smoked a day partcigs 18159 80.1 1 27 0.1 4493 Derived SMOKING variable


^îndicates 998)


^îndicates 999)

valid n

n % n % n % smoke 116 0.5 22564 6 category variables combining information from smoking and nofcigs. Values:0’never smoked’ 1’ex-smoker’ 2’occasional smoker’ 3’up to 10 a day’ 4’11 to 20 a day’ 5’more than 20 a day’.


SPSS SYNTAX missing values smoking othsmoke partcigs (8,9). missing values nofcigs (999). missing values agequit (99). compute smoke = -1. if (smoking = 1) smoke = 0. if (smoking = 2) smoke = 1. if (smoking = 3) smoke = 2. if (smoking = 4 and nofcigs <= 10) smoke = 3. if (smoking = 4 and (nofcigs > 10 and nofcigs <=20)) smoke = 4. if (smoking = 4 and nofcigs > 20) smoke = 5. missing values smoke (-1). variable labels smoke 'CM smoking habits'. value labels smoke 0'never smoked' 1'ex smoker' 2'occassional smoker' 3'up to 10 a day' 4'11 to 20 a day' 5'more than 20 a day'. frequencies variables=smoke.


DRINKING QUESTIONS variable system-missing user-missing: 8 (^

indicates 98; ^îndicates 998)


^îndicates 999)

valid n

n % n % n % drinks 66 0.3 7 27 0.1 22580 18260 (80.7%) CMs drank alcohol regularly (daily, weekly or monthly). 18260 asked how many of each type of alcoholic drink they had in last 7 days prior to interview beer 4420 19.5 ^^^16 0.1 18244 spirits 4420 19.5 ^^^1 ^^^16 0.1 18243 wine 4420 19.5 ^^^1 ^^^3 18256 sherry 4420 19.5 ^^^1 ^^^1 18258 pops 4420 19.5 ^^^1 ^^^1 18258 othdrink 4420 19.5 ^^^1 18259 cage1 493 2.2 5 22182 5521 (24.9%) CMs ever felt should cut down on their drinking. 5521 asked if felt this in last year cage2 17159 75.7 1 5520 cage3 493 2.2 22187 2193 (9.7%) CMs ever felt criticised about their drinking. 2193 asked if felt this in last year cage4 20487 90.3 2193 cage5 493 2.2 22187 2693 (11.9%) CMs ever felt bad or guilty about their drinking. 2693 asked if felt this in last year cage6 19987 88.1 2693 cage7 493 2.2 1 1 22185 874 (3.9%) CMs ever had a drink first thing in A.M. to steady hands. 874 asked if they had in last year cage8 21806 96.1 874 cage9 4208 18.6 4 0 18468 2539 (11.2%) CMs ever drink during breaks at work. 2539 asked how often they did this cage10 20141 88.8 2 2537 Derived DRINKING variables


^îndicates 998)


^îndicates 999)

valid n

n % n % n % drkunits 44250 18230 Continuous variable giving total number of alcohol units CM had in 7 days prior to interview. Combines information given in beer, spirits, wine, sherry and pops. Range from 0-427. Mean: 19.26, Median: 11.00, sd: 25.77 drkcut 499 2.2 22181 3 category variable combining cag1 and cage2. Values: 0'no' 1'yes, previously' 2'yes, last 12 months'. drkcrit 493 2.2 22187 3 category variable combining cag3 and cage4. Values: 0'no' 1'yes, previously' 2'yes, last 12 months'. drkbad 493 2.2 22187 3 category variable combining cag5 and cage6. Values: 0'no' 1'yes, previously' 2'yes, last 12 months'. drkhand 495 2.2 22185 3 category variable combining cag7 and cage8. Values: 0'no' 1'yes, previously' 2'yes, last 12 months'. drkwork 4214 18.6 18466 3 category variable combining cag9 and cage10. Values: 1'yes, special occasions' 2'yes, 2 or 3 times a month' 3'yes, 2 or 3 times a week' 4'yes, most days'.


SPSS SYNTAX missing values drinks (8,9). missing values beer spirits wine sherry pops (998,999). missing values othdrink (8). compute drkunits =((beer * 2) + spirits + wine + sherry + pops). variable labels drkunits 'CM alcohol units in a week'. frequencies variables=drkunits. missing values cage1 cage2 cage3 cage4 cage5 cage6 cage7 cage8 cage9 cage10 (9). compute drkcut = -1. if (cage1 = 2) drkcut = 0. if (cage1 = 1 and cage2 = 2) drkcut = 1. if (cage1 = 1 and cage2 = 1) drkcut = 2. missing values drkcut (-1). variable labels drkcut 'CM felt should cut down on drinking?'. compute drkcrit = -1. if (cage3 = 2) drkcrit = 0. if (cage3 = 1 and cage4 = 2) drkcrit = 1. if (cage3 = 1 and cage4 = 1) drkcrit = 2. missing values drkcrit (-1). variable labels drkcrit 'CM felt criticised about drinking?'. compute drkbad = -1. if (cage5 = 2) drkbad = 0. if (cage5 = 1 and cage6 = 2) drkbad = 1. if (cage5 = 1 and cage6 = 1) drkbad = 2. missing values drkbad (-1). variable labels drkbad 'CM felt bad/guilty about drinking?'. compute drkhand = -1. if (cage7 = 2) drkhand = 0. if (cage7 = 1 and cage8 = 2) drkhand = 1. if (cage7 = 1 and cage8 = 1) drkhand = 2. missing values drkhand (-1). variable labels drkhand 'CM had drink to steady hands?'. compute drkwork = -1. if (cage9 = 2) drkwork = 0. if (cage9 = 1 and cage10 = 4) drkwork = 1. if (cage9 = 1 and cage10 = 3) drkwork = 2. if (cage9 = 1 and cage10 = 2) drkwork = 3. if (cage9 = 1 and cage10 = 1) drkwork = 4. missing values drkwork (-1). variable labels drkwork 'CM had drink at work?'. value labels drkcut drkcrit drkbad drkhand 0'no' 1'yes, previously' 2'yes, last 12 months'. value labels drkwork 0'no' 1'yes, special occasions' 2'yes, 2 or 3 times a month' 3'yes, 2 or 3 times a week' 4'yes, most days'. frequencies variables=drkcut drkcrit drkbad drkhand drkwork .


ACCIDENT QUESTIONS variable system-missing user-missing: 98

(^ indicates 998) user-missing: 99 (^ indicates 999)

valid n

n % n % n % Multi-coded questions ask if CM had an accident, and if yes, what type of accident. accidan1 accidan2 accidan3 accidan4 accidan5 accidan6 accidan7 accidan8 accidan9

66 0.3 8 0 24 0.1 22582

12458 (55%) reported no accidents with 10124 (45%) reporting they had some kind of accident. These CMs asked how many accidents they had had in total, the age and type of their most recent accident accidno 12524 55.2 ^10 0.1 ^42 0.4 10104 accage 12576 55.4 ^1 0 ^18 0.2 10085 accwhy 12576 55.4 1 0 2 0 10101 Derived ACCIDENT variables



valid n

n % n % n % 2-category derived variables from multi-coded accidan1-accidan9. Each variable gives % experiencing type of accident: road accident (pedestrian), road accident (driver/passenger), at work, at home, at school/college, playing sport, other type, violent assault, sexual assault. Values: 0’not experienced’ 1’yes, experienced’. accroadp accroadd accwork acchome accschcl accsport accother vassault asssex

98 0.4 22582


SPSS SYNTAX missing values accidan1 (98,99). frequencies variables=accidan1 accidan2 accidan3 accidan4 accidan5 accidan6 accidan7 accidan8 accidan9 accidno accage accwhy. do if (accidan1 >= 0). compute accroadp = 0. compute accroadd = 0. compute accwork = 0. compute acchome = 0. compute accschcl = 0. compute accsport = 0. compute accother = 0. compute vassault = 0. compute asssex = 0. do repeat x = accidan1 accidan2 accidan3 accidan4 accidan5 accidan6 accidan7 accidan8 accidan9. if (x = 1) accroadp = 1. if (x = 2) accroadd = 1. if (x = 3) accwork = 1. if (x = 4) acchome = 1. if (x = 5) accschcl = 1. if (x = 6) accsport = 1. if (x = 7) accother = 1. if (x = 8) vassault = 1. if (x = 9) asssex = 1. end repeat. end if. variable labels accroadp 'CM experienced road accident as pedestrian'. variable labels accroadd 'CM experienced road accident as driver/passenger'. variable labels accwork 'CM experienced accident at work'. variable labels acchome 'CM experienced accident at home'. variable labels accschcl 'CM experienced accident at school/college'. variable labels accsport 'CM experienced accident when doing sport'. variable labels accother 'CM experienced other sort of accident'. variable labels vassault 'CM experienced violent assault or mugging'. variable labels asssex 'CM experienced sexual assault'. value labels accroadp accroadd accwork acchome accschcl accsport accother vassault asssex 0'not experienced' 1'yes, experienced'. frequencies variables=accroadp accroadd accwork acchome accschcl accsport accother vassault asssex .


APPENDIX 6: Problems with reporting of GCSEs Documentation produced by Andrew enkins of the Institute of Education relating to the reporting of GCSEs is reproduced below.


([email protected])

Report on GCSE date errors in NCDS/BCS 2000 (Andrew Jenkins) This report summarises the findings of a check on the dates at which cohort members in NCDS/BCS 2000 claim to have obtained their GCSEs. 1. The Extent of the Problem The earliest date at which it appears to be possible to have obtained a GCSE is 1988, as indicated by the following quotations: ‘GCSE was introduced in September 1986, with the first exams taking place in Summer 1988. It replaced GCE ‘O’ level and CSE’ (British Qualifications, 25 ed, 1994, Kogan Page). ‘Pupils leaving school in Summer 1988 sat either GCSEs or a mixture of ‘O’ levels and CSEs (and some sat all three types of exam)’. Dept of Education and Science, Statistics of School Examinations, 1990. ‘GCSE examinations were introduced in 1988. The final GCE ‘O’ levels examinations for home students were held in the winter session 1987, with summer 1987 being the final session for CSE exams’. Dept for Education, Statistics of School Examinations, 1991. Some 94 NCDS cohort members and over 1,000 BCS cohort members (CMs) are reporting having obtained one or more GCSEs before 1988 in the 2000 sweep of the data. All of the 94 NCDS CMs and the first 200 BCS CMs have been checked against earlier sweeps of the NCDS/BCS surveys. 2. GCSE Date Errors for the NCDS In the 2000 sweep of the data, those NCDS CMs who had participated in the previous sweep in 1991 were asked to supply information on all qualifications which they had obtained since 1991; those who were absent from the previous sweep were asked about all qualifications which they had obtained since 1974, or school leaving age. Asking for information over such a long spell of time is very likely to introduce inaccuracies due to imperfect recall. However, for NCDS, the 2000 sweep represents the sixth follow-up on the lives of cohort members which means that there is potentially a lot of information in earlier sweeps which we can use for checking. Back in 1978, schools and colleges attended by cohort members were contacted and asked to supply details of the qualifications which NCDS CMs had obtained. This is a valuable resource, since it represents an objective source of information on qualifications obtained. The main drawback is that, obviously, it only covers the period up to 1978. In what follows it is referred to as the EXAMs file. It contains data on the number of qualifications obtained, and also some data on the subjects of qualifications. NCDS IV (1981) appears to contain little or nothing on qualifications obtained, but NCDS V (1991) is more informative. In NCDS V CMs were asked (a) about qualifications ever obtained and (b) about qualifications obtained since 1981. Note that the NCDS questionnaire did ask for GCSEs, as well as O levels and CSEs. The main drawback of NCDS V is that it does not tell us about the number of qualifications.


For example, CMs were asked if they had CSEs at grades 2-5, but were not asked to report how many they had obtained. It also, of course, has problems of recall errors. The EXAMS file, supplemented by NCDS V, then, are the main sources which we can use to check the accuracy of the data in NCDS VI (2000). A detailed case-by-case listing for each of the 94 NCDS records in question is contained in Appendix 1, and this should be referred to when considering the summary given here. Fortunately, most of the 94 cases were present in the EXAMS file, and since the majority of erroneous cases were reporting that they got their GCSEs in the 1973 to 1976 period, the data contained in the EXAMS file is relevant. In more detail, all but ten of the 94 cases were present in the EXAMs file. The cases which were not in EXAMS were cases 2,16,26,45,62,74,80,83,91,93. Out of these, five (16,26,74,80,83) were present in NCDS 5. There is no data for the remaining five. For some of the remaining cases, EXAMS file data was found to be not relevant because they were reporting GCSE qualifications after 1978 but fortunately, this applied to only a small number of cases. These cases were (3,11,18,30,77,81,82). Some of these (3,30,82) were covered by NCDS V, but the remaining 4 were not. In sum, then, there were only 9 cases where data could not be obtained from earlier sweeps to check the GCSE errors in NCDS VI. The majority of those for whom data is available would appear to have CSEs rather than the GCSEs reported in NCDS VI. However, some have O levels, and some a mixture of O levels and CSEs. To give some specific examples, cases 13 and 15 are just two examples of a straightforward mix-up between GCSEs and CSEs (for case 13 there is also a small difference in number of GCSE/CSEs reported – 7 in the EXAMS file and 6 in 2000 sweep). In cases 9 and 37, on the other hand, GCSEs and O levels would appear to have been muddled up by the CMs concerned. Although a good many cases appear to show confusion between GCSEs and either O levels or CSEs, some cases are less simple. For instance, in case 94, there is a huge discrepancy between the number of O levels in the EXAMS file and the number of GCSEs listed in NCDS VI. Cases 19 and 52 are other cases where there is little evidence in the EXAMS file that the qualifications reported in NCDS VI were actually obtained. On the other hand, some cases can be easily sorted out even in the absence of data. Case 2 shows duplication and the GCSE data can be deleted. It is good that we have some data to check NCDS VI against for most of the 94 problem cases. Most of the GCSEs would appear to be CSEs, in fact, but some cases they are O levels or a mixture of O levels and CSEs, and in some cases it is not simple to decide what the best answer is. Nonetheless, it is probably best to deal with them on a case by case basis rather than adopting a simple but inaccurate solution such as simply recoding all GCSEs to O levels or all GCSEs to CSEs. 3. GCSE date errors among a sample of BCS 2000 CMs. Because there are such a large number of GCSE errors among the BCS cohort, it has only been possible to look at the first 200 of these (rather less than a fifth of the total). The BCS CMs were one of the last cohorts to sit O levels and CSEs as their school-leaving exams before they were replaced by GCSEs. BCS CMs should only have GCSEs if they stayed in education to at least age 18, or subsequently returned to study for a qualification. The only source of data for checking the responses in BCS 2000 is the earlier BCS survey conducted in 1996. Questions about qualifications in the BCS 1996 survey were asked in the order: CSE, O level, GCSE, rather than the order GCSE, O level, CSE in BCS 2000. Thus checking BCS 2000 against BCS 1996 should help to root out errors which may have arisen from the order in which the questions were asked in BCS 2000. However, BCS 1996 had a relatively small sample size (n = 9,003), so many of those in BCS 2000 may not be present in BCS 1996. Of the 200 BCS CMs checked, some 134 were also present in BCS 96. Out of these 134, 42 were also reporting GCSEs in BCS 96 (although not necessarily the same number as in BCS 2000); 17 provided no information about qualifications in BCS 96, and 75 did not report any GCSEs in BCS 96. Detailed information on each case is presented in Appendix 2.


Among the 134 for whom 1996 data are available it is usually possible to make some reasonable guess as to what should be done about the erroneous GCSE data in BCS 2000. For example, it is clear that in cases 8 and 11 the GCSE data should be recoded to O levels, while in case 41 the GCSEs should be recoded as CSEs. Other cases are not quite so obvious, and some, where there are large discrepancies between 1996 and 2000 (e.g. cases 23, 30) or conversely where errors have been repeated (e.g. 31) are fairly puzzling. The real difficulty is, however, that we have a substantial number of cases not present in 1996, and so there is no data to check the BCS 2000 results against. 4. O levels and CSEs after 1988 reported by CMs in NCDS/BCS 2000 I have also had a quick look at whether any NCDS cohort members are reporting having obtained O levels after they were (presumably) phased out in 1988. Among all the NCDS CMs, some 59 had reported one or more O levels with a date post 1988. The dates for the reported O levels were all in the 1990s. The source of the error is likely to be either that (a) they have actually got GCSEs rather than O levels or (b) a mistake was made inputting the data, so that the O level was actually obtained in,say, 1974 but this was coded as 1994. Now 47 of the 59 were also present in NCDS V, which means that they were (or should have been) only asked questions about qualifications obtained since 1991. This would make it unlikely that error (b) occurred. The data do indeed show that most of these 47 were only reporting O levels obtained in the 1990s: 28 reported a single O level obtained in the 1990s and no earlier O levels; a further 13 reported two O levels obtained in the 1990s and no O levels. However, the remaining six of the 47, despite being present in NCDS V in 1991, also reported O levels obtained in both the 1970s and the 1990s. For the 12 not present in 1991, it is difficult to discriminate between the (a) and (b) errors listed above. So it is not possible to resolve all the cases, but for many recoding to GCSE would be a sensible solution. Only two NCDS CMs reported CSEs after 1988. For BCS, there appear to be 133 cases in which CMs are reporting O levels after 1988, and 32 cases of CSEs after 1988 but these have not been studied in any detail.


Problems in Using the NCDS/BCS 2000 Data (Andrew Jenkins 27/3/2001) 1) The fact that there are three different sub-groups in the data, namely, the BCS whose qualifications are recorded from 1986; some NCDS who also participated in 1991 and so data on them is from 1991 onwards, and other NCDS who missed out last time and so have data in current sweep back to 1974 caused us some confusion. I think this will not be clear to other potential users and needs to be flagged very clearly in the documentation. 2) As you know, merging data with earlier sweeps is far from straightforward because of confusion about chrtid, and serial variables in the different sweeps. This has been a major headache for the QCA project where we needed to merge different sweeps. 3) Many of the qualification variables have 98 or 99 (or in some cases 8,9 or 9998, 9999) as value labels, but these are not defined. Presumably these are missing values/could not say/would not say, but they need to be defined. This problem is fairly minor for some variables e.g the year people got their CSEs (if they got them in a single year) = cseyrq and this has 6 people coded as having got their CSEs in either 9998 or 9999; but much more serious for other variables e.g ‘how many CSEs did you get at grade 1’ (edcse1) has 120 people coded as 99 and one coded as 98. Some cases (2,164 for this variable) are also coded as zero but this obviously doesn’t cover all the people who have not got a CSE. Why are some coded zero and not others? 4) Problem 3 also contributes to making it very difficult to get consistent answers to relatively simple questions such as how many people in the sample have got a CSE. One way to answer this is to use the derived variable numcse. This reports that 4,621 have a value for this answer and we can deduct the number reporting zero CSEs (251) to work out the number with one or more CSEs (4621-251= 4370). Another approach to the same question is to look at the variables edqtyp14 to edqtyp26 which report whether individual has got particular qualifications. If we use these variables to create a dummy variable, ncse, taking value 1 if individual has got CSE and zero if not, as follows compute ncse=0. if any(3,edqtyp14,edqtyp15,edqtyp16,edqtyp17,edqtyp18, edqtyp19,edqtyp20,edqtyp21,edqtyp22,edqtyp23,edqtyp24,edqtyp25, edqtyp26) ncse=1. execute . (note that 3 is the code for CSE for these variables). Doing this and running frequencies yields the answer that 4,621 have got (at least one) CSE. Thus we have two answers to the question how many have got one or more CSEs – is it 4,621 or 4,370? On closer inspection, it looks as if some of those reporting a qualification in the edqtyp group of variables have not in fact got it, so we have tended to use variables such as numcse, knocking out the zeros and 98s etc. But it would be good if we could get consistent answers to the question whichever way it was looked at. Note that this issue arises with the academic educational qualifications (edqtyp14 to edqtyp26) and with a similar set of vocational qualifications voctyp12 to voctyp22. 5) If the variable Dmpart=1 this implies that NCDS CM was included in 1991 sweep and therefore should not have pre 91 qualifications included on current sweep. However, there are individuals with Dmpart=1 and qualifications recorded with pre 91 dates. For example, with A levels, if we take edyear56, the year in which people got their first A level, then 105 of the 179 individuals with an A


level in NCDS and with Dmpart=1 are recorded as having got their A level before 1991 (mostly in 1976). This problem also arises with other qualifications. Is this because some individuals who did participate in 1991 have nonetheless had their qualifications recorded again in 2000 or is the Dmpart variable faulty in that it is misallocating significant numbers of people (my money’s on the former, I think). 6) As already mentioned, large nos of the BCS people are claiming to have obtained GCSEs in 1986 when they were 16. My impression is that GCSE was not introduced until about 1988 and that many respondents have just got mixed up with O levels and CSEs. Conversely, some people are reported as having got CSEs and O levels after the late 1980s when presumably they had been abolished. 7) Proxy individuals have also caused a little confusion since the variables in which they report qualifications have very similar names to the variable names for the rest of the sample. Again this is probably just be a case of flagging very clearly that they are there in the documentation. 8) There are a couple of typos, as well: Labelling of the variable eddiped incorrect. Should be ‘Does CM know the year s/he got their Dip H Ed’ not ‘Year obtained Dip H Ed’. The variable edgcse2 is labelled as ‘number of GCSEs obtained at grades D-E’; should be ‘number of GCSEs obtained at grades D-G’, (?) as with other qualifications such as AS levels.


Summary of Results of Checking Dates at which CM’s obtained GCSEs in NCDS/BCS 2000. Earliest Dates at which a GCSE could have been obtained: The earliest date at which it appears to be possible to have obtained a GCSE is 1988 (see the quotations below). Some 94 NCDS cohort members and over 1,000 BCS cohort members are reporting having obtained one or more GCSEs before 1988 in the 2000 sweep of the data. ‘GCSE was introduced in September 1986, with the first exams taking place in Summer 1988. It replaced GCE ‘O’ level and CSE’ (British Qualifications, 25 ed, 1994, Kogan Page). ‘Pupils leaving school in Summer 1988 sat either GCSEs or a mixture of ‘O’ levels and CSEs (and some sat all three types of exam)’. Dept of Education and Science, Statistics of School Examinations, 1990. ‘GCSE examinations were introduced in 1988. The final GCE ‘O’ levels examinations for home students were held in the winter session 1987, with summer 1987 being the final session for CSE exams’. Dept for Education, Statistics of School Examinations, 1991. NCDS For the 94 NCDS cohort members examing NCDS 5 and the EXAMS file from 1978 provided information on the qualifications obtained by all but nine of the 94. The following were not in the EXAMS file: cases 2,16,26,45,62,74,80,83,91,93. Some of these (16,26,74,80,83) were present in NCDS 5. There is no data for the remaining five. For some of the remaining cases, EXAMS file data was found to be not relevant because they were reporting GCSE qualifications after 1978. These cases were (3,11,18,30,77,81,82). Some of these (3,30,82) were covered by NCDS 5, but the remaining 4 were not. In total, then, there were only 9 cases where data could not be obtained from earlier sweeps to check the GCSE errors in NCDS 6. The majority of those for whom data is available would appear to have CSEs rather than the GCSEs reported in NCDS 6. However, some have O levels, and some a mixture of O levels and CSEs. BCS I have looked at the first 200 cases for BCS. Of these, some 134 were present in BCS 96. Out of these 134, 42 were also reporting GCSEs in BCS 96 (although not necessarily the same number as in BCS 2000); 17 provided no information about qualifications in BCS 96, and 75 did not report any GCSEs in BCS 96.


NCDS VI – Results of Checking those with Erroneous Dates for GCSEs against Earlier NCDS Sweeps There are 94 cases in NCDS VI in which NCDS cohort members have recorded a date for their GCSEs prior to the introduction of the first GCSE exams (1988). These 94 cases are as follows. Case: 1 Chrtid: 581058E NCDS VI GCSE/CSE/GCE Qualifications: one GCSE in 1974, also 3 O levels in 1974. Present in NCDS V: Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 3 CSEs at grades 2-5. Case: 2 Chrtid: 330031E NCDS VI GCSE/CSE/GCE Qualifications: GCSEs from 1975 in homecrafts, English literature and English language. Also CSEs from 1976 in English Language, English Literature, Geography, Drama and Homecraft. Also O levels from 1977 in English Language, English Literature, Geography, Drama and Religious Knowledge. Present in NCDS V: No Present in EXAMS file: No EXAMS file GCE/CSE Qualifications: N/A. Case: 3 Chrtid: 750128Q NCDS VI GCSE/CSE/GCE Qualifications: 3 GCSEs in 1982, 2 at A-C grades, 1 at lower grade. No O levels or CSEs. Present in NCDS V: Yes NCDS V Qualifications: Recorded as having ‘other’ qualification(s), but not GCSE, CSE or O level. Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: No qualifications. Case: 4 Chrtid: 845002U NCDS VI GCSE/CSE/GCE Qualifications: 8 GCSEs in 1973, 4 at A-C grades and 4 at lower grades. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 O level/CSE grade 1, 2 CSEs at grade 4, 2 CSEs at grade 5. Case: 5 Chrtid: 516081F NCDS VI GCSE/CSE/GCE Qualifications: 5 GCSEs in 1974, 4 at A-C grades, 1 at lower grade. No O levels or CSEs. Present in NCDS V: Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE grade 4, 2 CSE grade 5. Case: 6 Chrtid: Y30263X NCDS VI GCSE/CSE/GCE Qualifications: 5 GCSEs in 1974, 3 at A-C grades, 2 at lower grades. 2 O levels in 1974, no CSEs. Present in NCDS V:


Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 O level/CSE grade 1; 2 CSE grade 4. Case: 7 Chrtid: Y30072Q NCDS VI GCSE/CSE/GCE Qualifications: 6 GCSEs all at A-C grades in 1974. No O levels or CSEs. Present in NCDS V: Present in EXAMS file: EXAMS file GCE/CSE Qualifications: 2 CSEs at grade 3; 2 CSEs at grade 4, 2 CSEs at grade 5. Case:8 Chrtid:X66014T NCDS VI GCSE/CSE/GCE Qualifications:3 GCSEs in 1974, no O levels or CSEs. Present in NCDS V:Yes Present in EXAMS file:yes EXAMS file GCE/CSE Qualifications:1 O level at grade A-C. Case:9 Chrtid:X83029D NCDS VI GCSE/CSE/GCE Qualifications:7 GCSEs in 1974, all at A-C grades, no O levels or CSEs. Present in NCDS V: No Present in EXAMS file:Yes EXAMS file GCE/CSE Qualifications:7 O levels at A-C, no CSEs. Case:10 Chrtid:048054V NCDS VI GCSE/CSE/GCE Qualifications:6 GCSEs in 1974, no O levels or CSEs. Present in NCDS V: No Present in EXAMS file:Yes EXAMS file GCE/CSE Qualifications:1 CSE grade 5. Case:11 Chrtid:044015N NCDS VI GCSE/CSE/GCE Qualifications:4 GCSEs – 1 obtained in 1980, 1 in 1994, 2 in 1995 Present in NCDS V: No Present in EXAMS file:Yes EXAMS file GCE/CSE Qualifications: No O levels or CSEs. Case: 12 Chrtid: 620094X NCDS VI GCSE/CSE/GCE Qualifications: 2 GCSEs in 1974, both at A-C grades, no O levels and no CSEs. Present in NCDS V: No Present in EXAMS file: EXAMS file GCE/CSE Qualifications: 2 CSEs, 1 at grade 2; 1 at grade 4. Case: 13 Chrtid:100052D NCDS VI GCSE/CSE/GCE Qualifications: 6 GCSEs in 1974, all at A-C grades Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 3 CSE grade 2; 3 CSE grade 3; 1 CSE grade 5.


Case:14 Chrtid: 517088B NCDS VI GCSE/CSE/GCE Qualifications: 3 GCSEs in 1974, 2 at A-C grades and 1 at D-G. Also 1 CSE grade 1 in 1973. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: Took 3 exams in total; passed two: 1 CSE grade 3, 1 CSE grade 4. Case:15 Chrtid: 081020Q NCDS VI GCSE/CSE/GCE Qualifications: 1 GCSE in 1974 at A-C grade, no O levels and no CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE at grade 2. Case: 16 Chrtid: 835005U NCDS VI GCSE/CSE/GCE Qualifications: 10 GCSEs in 1974, all at grades A-C. Present in NCDS V: Yes Qualifications in NCDS V: CSEs at grade 1, and CSEs at grades 2-5. Present in EXAMS file: No EXAMS file GCE/CSE Qualifications: N/A. Case: 17 Chrtid: 093052W NCDS VI GCSE/CSE/GCE Qualifications: 3 GCSEs in 1973, 2 of them at A-C grade, and 1 at a lower grade; 1 O level in 1973 at A-C grade; 3 CSEs in 1973, 2 at grade 1, and 1 at a lower grade. Present in NCDS V: Yes. Qualifications in NCDS V: CSEs at grade 1, and O levels at A-C grades. Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE grade 1, 1 CSE grade 4, 2 CSE grade 5. Case: 18 Chrtid: 093070Y NCDS VI GCSE/CSE/GCE Qualifications: 1 GCSE in 1986; 5 O levels in 1974 Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 6 CSEs at grades 2-5: 1 at grade 2, 2 at grade 3, 1 at grade 4, 2 at grade 5. Case: 19 Chrtid: 0930175P NCDS VI GCSE/CSE/GCE Qualifications: 6 GCSEs in 1974, all at below grade C. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: No exams taken or passed. Case: 20


Chrtid: 222005A NCDS VI GCSE/CSE/GCE Qualifications: 4 GCSEs in 1974 at A-C; 4 CSEs in 1974 at grade 1. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE at grade 4. Case: 21 Chrtid: 054006T NCDS VI GCSE/CSE/GCE Qualifications: 4 GCSEs in 1974, 3 at A-C grades, 1 at a lower grade. No O levels or CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE at grade 2, 3 CSEs at grade 5. Case: 22 Chrtid: 055031X NCDS VI GCSE/CSE/GCE Qualifications: 4 GCSEs in 1974, all at A-C grades; no O levels or CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 4 exams taken, 3 passed: 2 CSE grade 3, 1 CSE grade 4 Case: 23 Chrtid: 093071A NCDS VI GCSE/CSE/GCE Qualifications: 2 GCSEs in 1974, 1 at grade A-C; no O levels, no CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE grade 1; 1 CSE grade 4. Case: 24 Chrtid: 083047X NCDS VI GCSE/CSE/GCE Qualifications: 1 GCSE in 1976 at A-C grade; 6 CSEs in 1974, 3 at grade 1, 3 at grades 2-5. Present in NCDS V: No Present in EXAMS file: EXAMS file GCE/CSE Qualifications: 6 CSEs: 3 at grade 1, 1 at grade 2, 1 at grade 3, 1 at grade 5. Case: 25 Chrtid: 099010M NCDS VI GCSE/CSE/GCE Qualifications: 5 GCSEs in 1974, 4 at A-C grades, 1 at lower grades. No O levels and no CSEs. Present in NCDS V: Yes Qualifications in NCDS V: City & Guilds qualifications only. Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: No O levels or CSEs obtained. Case: 26 Chrtid: 094022S NCDS VI GCSE/CSE/GCE Qualifications: 5 GCSEs in 1974 (3 at A-C grades, 2 at lower grades). No O levels, 3 CSEs in 1974, all at grades 2-5. Present in NCDS V: Yes Qualifications in NCDS V: CSEs at grades 2-5 only.


Present in EXAMS file: No EXAMS file GCE/CSE Qualifications: N/A. Case: 27 Chrtid: 055110T NCDS VI GCSE/CSE/GCE Qualifications: 8 GCSEs in 1976, 5 at A-C grades, 3 at lower grades. No O levels or CSEs. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 4 O level/CSE grade 1 passes. 1 CSE grade 2, 1 CSE grade 3. Case: 28 Chrtid: 092153X NCDS VI GCSE/CSE/GCE Qualifications: 1 GCSE in 1974 at D-G grade. Also 5 CSEs all at grade 1. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 4 CSEs at grade 1; 1 CSE grade 3. Case: 29 Chrtid: 984038V NCDS VI GCSE/CSE/GCE Qualifications: 4 GCSEs in 1974, 2 at A-C grades, 2 at D-G grades. No O levels or CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE grade 2, 1 CSE grade 4, 1 CSE grade 5. Case: 30 Chrtid: Y30042E NCDS VI GCSE/CSE/GCE Qualifications: 3 GCSEs at D-G in 1985. No O levels or CSEs. Present in NCDS V: Yes Qualifications in NCDS V: CSEs at grades 2-5; O levels at A-C grades; also RSA and City & Guilds qualifications. Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE grade 5. Case: 31 Chrtid: 282068R NCDS VI GCSE/CSE/GCE Qualifications: 2 GCSEs in 1974, 3 more in 1985: 3 at A-C grades, 2 at D-G grades. No O levels or CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE grade 1, 1 CSE grade 3, 1 CSE grade 4. Case: 32 Chrtid: 110315S NCDS VI GCSE/CSE/GCE Qualifications: 2 GCSEs at A-C grades, 1 in 1974, the other in 1977; no O levels or CSEs. Present in NCDS V: Yes


Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE grade 2, 4 CSEs at grade 4. Case: 33 Chrtid: 330002X NCDS VI GCSE/CSE/GCE Qualifications: 1 GCSE in 1977 at A-C grade; also 5 CSEs in 1974, 1 at grade 1, and the other 4 at grades 2-5. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE grade 1, 3 CSEs at grade 3, 1 CSE grade 5. Case: 34 Chrtid: 330092B NCDS VI GCSE/CSE/GCE Qualifications: 5 GCSEs in 1974 all at A-C grades; no O levels or CSEs. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 2 CSEs at grade 3, 2 CSEs at grade 4, 1 CSE at grade 5. Case: 35 Chrtid: 330009N NCDS VI GCSE/CSE/GCE Qualifications: 4 GCSEs in 1974, 2 of these at A-C grades, 2 at D-G. 1 O level in 1974 at grades D-E, no CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: No O levels recorded; 1 CSE grade 3, 1 CSE grade 4, 2 CSEs at grade 5. Case: 36 Chrtid: 330074Z NCDS VI GCSE/CSE/GCE Qualifications: 5 GCSEs in 1974, all at grades D-G; also 2 O levels in 1974 at A-C grades; no CSEs. Present in NCDS V: Yes Qualifications in NCDS V: O levels at A-C grades. Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE grade 3, 1 CSE grade 5. Case: 37 Chrtid: 382013W NCDS VI GCSE/CSE/GCE Qualifications: 2 GCSEs in 1974, both at A-C grades; also 5 CSEs in 1974, all at grades 2-5. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 2 O levels at grades A-C; 1 CSE grade 3, 1 CSE grade 4, 2 CSE grade 5.


Case: 38 Chrtid: 238018Z NCDS VI GCSE/CSE/GCE Qualifications: 6 GCSEs in 1974; 4 at A-C grades, 2 at D-G grades. No O levels or CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 2 CSE grade 2, 3 CSE grade 3, 2 CSE grade 4. Case: 39 Chrtid: Y20012P NCDS VI GCSE/CSE/GCE Qualifications: 7 GCSEs in 1976, 5 at A-C grades, 2 at D-G. No O levels and no CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE grade 2, 3 CSE grade 3, 3 CSE grade 4. Case: 40 Chrtid: 110106H NCDS VI GCSE/CSE/GCE Qualifications: 5 GCSEs in 1974, all at A-C grades. No O levels or CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE grade 3, 1 CSE grade 4, 1 CSE grade 5. Case: 41 Chrtid: 120011B NCDS VI GCSE/CSE/GCE Qualifications: 3 GCSEs in 1974, all at A-C. No O levels or CSEs. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE grade 2, 1 CSE grade 3, 2 CSE grade 4, 1 CSE grade 5. Case: 42 Chrtid: 052018Q NCDS VI GCSE/CSE/GCE Qualifications: 1 GCSE in 1975, 1 in 1996, both at A-C grades. No O levels and no CSEs. Present in NCDS V: Yes Qualifications in NCDS V: O levels at A-C grades. Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: No exams taken or passed. Case: 43 Chrtid: 184008Z NCDS VI GCSE/CSE/GCE Qualifications: 3 GCSEs in 1974, 2 at A-C grades, 1 at D-G. Present in NCDS V: Yes Qualifications in NCDS V: CSEs at grades 2-5. Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: No exams taken or passed.


Case: 44 Chrtid: 937018X NCDS VI GCSE/CSE/GCE Qualifications: 5 GCSEs in 1974, all at A-C grades. No O levels and no CSEs. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 2 O level/CSE grade 1, 1 CSE grade 3, 2 CSE grade 5. Case: 45 Chrtid: 935018M NCDS VI GCSE/CSE/GCE Qualifications: 1 GCSE in 1974 at A-C grade; no O levels and no CSEs. Present in NCDS V: No Present in EXAMS file: No EXAMS file GCE/CSE Qualifications: N/A. Case: 46 Chrtid: Y30070L NCDS VI GCSE/CSE/GCE Qualifications: 4 GCSEs in 1974, all at A-C grades. No O levels or CSEs. Present in NCDS V: Yes Qualifications in NCDS V: CSEs at grades 2-5. Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: No exams taken or passed. Case: 47 Chrtid: 937014P NCDS VI GCSE/CSE/GCE Qualifications: 3 GCSEs in 1974, all at A-C grades. No O levels and no CSEs. Present in NCDS V: Yes Qualifications in NCDS V: Recorded as having no qualifications. Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: No exams taken or passed. Case: 48 Chrtid: 381033X NCDS VI GCSE/CSE/GCE Qualifications: 6 GCSEs in 1974, 1 at A-C grade, 5 at D-G grades. No O levels or CSEs. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE grade 3, 2 CSE grade 4, 3 CSE grade 5. Case: 49 Chrtid: 381064K NCDS VI GCSE/CSE/GCE Qualifications: 3 GCSEs in 1976, all at A-C grades. No O levels, 3 CSE grade 1 also in 1976. Present in NCDS V: Yes Present in EXAMS file:Yes EXAMS file GCE/CSE Qualifications: 4 CSEs at grades 2-5; 1 at grade 2, 1 at grade 3, 1 at grade 4, 1 at grade 5.


Case: 50 Chrtid: 282113T NCDS VI GCSE/CSE/GCE Qualifications: 5 GCSEs in 1973, 1 at an unspecified date, all at A-C grades. No O levels, no CSEs. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 O level/CSE grade 1, 2 CSE grade 2, 1 CSE grade 3. Case: 51 Chrtid: 385059P NCDS VI GCSE/CSE/GCE Qualifications: 13 GCSEs, 10 in 1974, 2 in 1975, 1 in 1995; 12 at A-C grades, 1 at D-G. No O levels, and no CSEs. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 8 O levels at A-C grades; no CSEs. Case: 52 Chrtid: 092346K NCDS VI GCSE/CSE/GCE Qualifications: 2 GCSEs in 1974, both at A-C grades. No O levels and no CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: No Exams taken or passed. Case: 53 Chrtid: 527049X NCDS VI GCSE/CSE/GCE Qualifications: 7 GCSEs in 1974, all at A-C grades. No O levels and no CSEs. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 7 exams taken, 6 passes. 5 CSEs at grade 3; 1 CSE at grade 4. Case: 54 Chrtid: 825099A NCDS VI GCSE/CSE/GCE Qualifications: 2 GCSEs in 1974, 2 in 1985. 2 of these were at A-C grades, 2 at lower grades. No O levels or CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 3 O level/CSE 1. 2 at CSE grade 2. Case: 55 Chrtid: 982023V NCDS VI GCSE/CSE/GCE Qualifications: 3 GCSEs in 1974, all at A-C grades. No O levels or CSEs. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 2 CSE grade 4, 1 CSE grade 5. Case: 56 Chrtid: 982081L NCDS VI GCSE/CSE/GCE Qualifications: 1 GCSE in 1974, A-C grade. 3 CSEs at grade 1, 2 CSEs at grades 2-5. Present in NCDS V: No


Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 O level/CSE 1; 1 CSE grade 2, 1 CSE grade 3, 1 CSE grade 4, 1 CSE grade 5. Case: 57 Chrtid: 960080K NCDS VI GCSE/CSE/GCE Qualifications: 4 GCSEs in 1974, all at A-C grades; 1 O level in 1974, no CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE grade 1, 2 CSE grade 3, 3 CSE grade 4. Case: 58 Chrtid: Y30017F NCDS VI GCSE/CSE/GCE Qualifications: 8 GCSEs in 1974, all at A-C grades; no O levels and no CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 8 O levels, no CSEs listed. Case: 59 Chrtid: 509226R NCDS VI GCSE/CSE/GCE Qualifications: 4 GCSEs in 1976, all at A-C grades; no O levels; 7 CSEs in 1974, all at grades 2-5. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 3 O level/CSE grade 1, 2 CSEs at grade 2, 4 CSEs at grade 3, 1 CSE at grade 4. Case: 60 Chrtid: 511141V NCDS VI GCSE/CSE/GCE Qualifications: 2 GCSEs in 1973, both at A-C grades. No O levels reported, 6 CSEs in 1974 at grades 2-5. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 O level/CSE grade 1, 1 CSE grade 2, 2 CSE grade 3, 4 CSE grade 5. Case: 61 Chrtid: Y30098K NCDS VI GCSE/CSE/GCE Qualifications: 5 GCSEs in 1974, all at D-G grades; no O levels, 3 CSEs in 1974, all at grades 2-5. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE at grade 3, 2 CSEs at grade 5. Case: 62 Chrtid: 520034V NCDS VI GCSE/CSE/GCE Qualifications: 3 GCSEs in 1974, 1 at A-C grades, 2 at lower grades. No O levels or CSEs. Present in NCDS V: No Present in EXAMS file: No EXAMS file GCE/CSE Qualifications: N/A


Case: 63 Chrtid: 815061S NCDS VI GCSE/CSE/GCE Qualifications: 12 GCSEs, 11 obtained in 1974 and 1 in 1975, 6 at A-C grades and 6 at lower grades. No O levels or CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 12 exams taken; 9 passes at O level/CSE grade 1. Case: 64 Chrtid: 500421Q NCDS VI GCSE/CSE/GCE Qualifications: 4 GCSEs in 1973, all at A-C grades; 2 O levels, 1 at A-C grades, 1 at a lower grade; no CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE at grade 3, 1 CSE at grade 4, 1 CSE at grade 5. Case: 65 Chrtid: 684015K NCDS VI GCSE/CSE/GCE Qualifications: 6 GCSEs in 1974, 1 at A-C grade, the rest at lower grades; no O levels or CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: No exams taken, none passed Case: 66 Chrtid: 180025C NCDS VI GCSE/CSE/GCE Qualifications: 8 GCSEs in 1974, 4 at A-C grades, 4 lower. No O levels and no CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 3 O level/CSE grade 1. Case: 67 Chrtid: 528005F NCDS VI GCSE/CSE/GCE Qualifications: 5 GCSEs in 1974, 4 at A-C, 1 at a lower grade. No O levels and no CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 2 O level/CSE 1. Case: 68 Chrtid: Y30061K NCDS VI GCSE/CSE/GCE Qualifications: 2 GCSEs in 1974, both at A-C grades. No O levels and no CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE grade 4, 1 CSE grade 5. Case: 69 Chrtid: 610020K NCDS VI GCSE/CSE/GCE Qualifications: 3 GCSEs in 1976, 2 in 1974, all at A-C grades. 1 CSE grade 1, 2 CSEs at grades 2-5. Present in NCDS V: Yes


Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 O level/CSE grade 1, 2 CSEs at grade 2, 2 CSEs at grade 3, 1 CSE grade 4. Case: 70 Chrtid: 583014U NCDS VI GCSE/CSE/GCE Qualifications: 5 GCSEs in 1977, all at A-C grades. No O levels, 4 CSEs at below grade 1. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE at grade 5. Case: 71 Chrtid: 682037K NCDS VI GCSE/CSE/GCE Qualifications: 2 GCSEs in 1974, both at A-C grades. No O levels or CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 2 O level/CSE grade 1, 4 CSEs at grade 3, 1 CSE grade 5. Case: 72 Chrtid: 782118S NCDS VI GCSE/CSE/GCE Qualifications: 4 GCSEs in 1975, all at A-C grades. No O levels or CSEs. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE grade 3, 1 CSE grade 4. Case: 73 Chrtid: 782170U NCDS VI GCSE/CSE/GCE Qualifications: 8 GCSEs in 1974, 5 at A-C grades, 3 at lower grades. Also 2 O levels at A-C grades, no CSEs. Present in NCDS V: Yes Qualifications in NCDS V: CSE at grades 2-5, O levels at A-C. Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: No exams taken, no qualifications. Case: 74 Chrtid: 823501Q NCDS VI GCSE/CSE/GCE Qualifications: 3 GCSEs in 1974, all at A-C grades. No O levels or CSEs. Present in NCDS V: Yes Qualifications in NCDS V: CSEs at grade 1 and CSEs at grades 2-5. Present in EXAMS file: No EXAMS file GCE/CSE Qualifications: N/A Case: 75 Chrtid: 840002S NCDS VI GCSE/CSE/GCE Qualifications: 13 GCSEs in 1975, all at A-C grades. No O levels or CSEs. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 13 exams taken, 8 passes at O level/CSE grade 1. Case: 76 Chrtid: 840017F


NCDS VI GCSE/CSE/GCE Qualifications: 13 GCSEs in 1974, 10 at A-C grades, 3 at lower grades. No O levels or CSEs. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 9 O levels/CSE grade 1. Case: 77 Chrtid: 560023R NCDS VI GCSE/CSE/GCE Qualifications: 4 GCSEs , 2 in 1986, 2 in 1987, all at A-C grades. No O levels or CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 3 CSEs at grade 3. Case: 78 Chrtid: 882083H NCDS VI GCSE/CSE/GCE Qualifications: 9 GCSEs in 1974, all at A-C grades. No O levels or CSEs. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 8 O levels/CSE grade 1. Case: 79 Chrtid: 824514E NCDS VI GCSE/CSE/GCE Qualifications: 4 GCSEs in 1975, all at lower grades. No O levels or CSEs. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE at grade 2, 1 CSE at grade 3, 1 CSE at grade 4, 2 CSEs at grade 5. Case: 80 Chrtid: 882031M NCDS VI GCSE/CSE/GCE Qualifications: 5 GCSEs in 1974, all at A-C grades. No O levels. 1 CSE grade 1 in 1974. Present in NCDS V: Yes Qualifications in NCDS V: CSEs at grades 2-5; O levels at grades A-C; GCSEs at grades A-C. Present in EXAMS file: No EXAMS file GCE/CSE Qualifications: N/A. Case: 81 Chrtid: 365003W NCDS VI GCSE/CSE/GCE Qualifications: 2 GCSEs, 1 in 1986, 1 in 1996, both at A-C grades. 8 O levels, all at A-C grades, no CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 7 O levels/CSE grade 1. Case: 82 Chrtid: 610012L NCDS VI GCSE/CSE/GCE Qualifications: 1 GCSE in 1985, at A-C grade. No O levels or CSEs. Present in NCDS V: Yes Qualifications in NCDS V: CSEs at grade 1; O levels at grades A-C. Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE grade 3, 1 CSE grade 4, 1 CSE grade 5. Case: 83 Chrtid: 550237E NCDS VI GCSE/CSE/GCE Qualifications: 3 GCSEs in 1974, all at A-C grades. No O levels or CSEs.


Present in NCDS V: Yes Qualifications in NCDS V: CSEs at grades 2-5. Present in EXAMS file: No EXAMS file GCE/CSE Qualifications: N/A Case: 84 Chrtid: 933007V NCDS VI GCSE/CSE/GCE Qualifications: 7 GCSEs in 1974, all at A-C grades. No O levels or CSEs. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 3 O level/CSE grade 1, 2 CSE grade 3, 1 CSE grade 4. Case: 85 Chrtid: Y01195T NCDS VI GCSE/CSE/GCE Qualifications: 6 GCSEs in 1974, 1 at A-C grades, 5 at lower grades. No O levels or CSEs. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE at grade 2, 2 CSEs at grade 3, 1 CSE at grade 4, 2 CSEs at grade 5. Case: 86 Chrtid: 517102S NCDS VI GCSE/CSE/GCE Qualifications: 4 GCSEs in 1974, all at A-C grades. No O levels or CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE grade 4, 1 CSE grade 5. Case: 87 Chrtid: 500073T NCDS VI GCSE/CSE/GCE Qualifications: 8 GCSEs in 1974, all at A-C grades. No O levels or CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 3 CSE at grade 3, 3 CSE at grade 4. Case: 88 Chrtid: 825035X NCDS VI GCSE/CSE/GCE Qualifications: 4 GCSEs, 3 in 1986 and 1 in 1975, 1 is at A-C grade, others at lower grade. No O levels, 6 CSEs in 1975, 1 at grade 1, 5 at grades 2-5. Present in NCDS V: Yes Qualifications in NCDS V: CSEs at grades 2-5; CSEs at grade 1, O levels at A-C grade. Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 O level/CSE grade 1, 1 CSE grade 2, 3 CSE grade 3, 1 CSE grade 4, 1 CSE grade 5. Case: 89 Chrtid: 180007A NCDS VI GCSE/CSE/GCE Qualifications: 1 GCSE in 1973, at below grade C. No O levels, 5 CSEs in 1974 at grades 2-5. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: No exams taken, no qualifications.


Case: 90 Chrtid: 465013H NCDS VI GCSE/CSE/GCE Qualifications: 1 GCSE in 1974, at A-C grade. No O levels or CSEs. Present in NCDS V: No Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: No exams taken, no qualifications. Case: 91 Chrtid: 592014V NCDS VI GCSE/CSE/GCE Qualifications: 7 GCSEs in 1974, 5 at A-C grades, 2 at lower grades. No O levels or CSEs. Present in NCDS V: No Present in EXAMS file: No EXAMS file GCE/CSE Qualifications: N/A. Case: 92 Chrtid: 514058A NCDS VI GCSE/CSE/GCE Qualifications: 4 GCSEs in 1974, all at A-C grades. No O levels or CSEs. Present in NCDS V: Yes Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 CSE at grade 2, 2 CSEs at grade 3, 3 CSEs at grade 4. Case: 93 Chrtid: 514071S NCDS VI GCSE/CSE/GCE Qualifications: 6 GCSEs in 1974, all at A-C grades. 1 O level in 1974, no CSEs. Present in NCDS V: No Present in EXAMS file: No EXAMS file GCE/CSE Qualifications: N/A Case: 94 Chrtid: 503018A NCDS VI GCSE/CSE/GCE Qualifications: 12 GCSEs in 1973, 6 at A-C grades, 6 at lower grades. No O levels and no CSEs. Present in NCDS V: Yes Qualifications in NCDS V: O levels at A-C grades. Present in EXAMS file: Yes EXAMS file GCE/CSE Qualifications: 1 O level/CSE grade 1.


BCS70 1. CHRTID/SERIAL: 17347090 BCS 2000 Qualifications: 6 GCSEs, all in 1986. No O levels or CSEs. Present in BCS 1996: No 2. CHRTID/SERIAL: 17345113 BCS 2000 Qualifications: 5 GCSEs, all in 1986, 4 at grades A-C, 1 at a lower grade. No O levels or CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 1 O level at grade A-C; 7 GCSEs all at A-C grades. 3. CHRTID/SERIAL: 10482048 BCS 2000 Qualifications: 1 GCSE in 1987, grade A-C; 6 (or 8 as suggested by summary data) O levels in 1986; no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 8 O levels at grade A-C; also 1 O level at a lower grade. 1 GCSE. 4. CHRTID/SERIAL: 7246058 BCS 2000 Qualifications: 3 GCSEs in 1986, all at A-C grade. No O levels or CSEs. Present in BCS 1996: No 5. CHRTID/SERIAL: 5078026 BCS 2000 Qualifications: 2 GCSEs in 1986, both at A-C grades; no O levels; 7 CSEs in 1986. Present in BCS 1996: Yes. BCS 1996 Qualifications: 1 O level at A-C grade; 2 other O levels at lower grades; 1 CSE grade 1; 4 CSEs at below grade 1. No GCSEs. 6. CHRTID/SERIAL: 3535025 BCS 2000 Qualifications: 6 GCSEs in 1986, 4 at A-C grade, 2 at lower grades. No O levels or CSEs. Present in BCS 1996: No 7. CHRTID/SERIAL: 5110044 BCS 2000 Qualifications: 5 GCSEs in 1986, 3 at A-C grades, 2 at lower grades. Present in BCS 1996: No 8. CHRTID/SERIAL: 5840013 BCS 2000 Qualifications: 10 GCSEs in 1985. 10 O levels – 1 in 1983, 1 in 1984, 2 in 1985, 6 in 1986. No CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 10 O levels, all at A-C grades, and 5 GCSEs at A-C grades. 9. CHRTID/SERIAL: 5045072 BCS 2000 Qualifications: 5 GCSEs in 1986. No O levels or CSEs. Present in BCS 1996: No 10. CHRTID/SERIAL: 12072019 BCS 2000 Qualifications: 1 GCSE in 1987. 7 O levels - 3 in 1987 and 4 in 1986. 2 CSEs in 1986. Present in BCS 1996: No 11. CHRTID/SERIAL: 6266032 BCS 2000 Qualifications: 6 GCSEs in 1986, 5 at A-C grades, 1 at lower grade. No O levels or CSEs. Present in BCS 1996: Yes. BCS 1996 Qualifications: 5 O levels at A-C grades; no GCSEs and no CSEs.


12. CHRTID/SERIAL: 6287060 BCS 2000 Qualifications: 4 GCSEs in 1986, all at A-C grades; no O levels or CSEs. Present in BCS 1996: No 13. CHRTID/SERIAL: 21453000 BCS 2000 Qualifications: 5 GCSEs in 1987, 3 at A-C grades, 2 at lower grades. No O levels or CSEs. Present in BCS 1996: No 14. CHRTID/SERIAL: 3159070 BCS 2000 Qualifications: 1 GCSE in 1986, at A-C grade. 9 O levels, 8 in 1986 and 1 in 1985, all at A-C grades. No CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 10 O levels at A-C grades; no GCSEs. 15. CHRTID/SERIAL: 4539004 BCS 2000 Qualifications: 1 GCSE in 1986, at A-C grade. 3 O levels in 1986, at A-C grades. 3 CSEs in 1986, 2 at grade 1, and 1 at lower grade. Present in BCS 1996: No 16. CHRTID/SERIAL: 17549095 BCS 2000 Qualifications: 8 GCSEs in 1986, 7 at A-C grades and 1 at lower grade. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 7 O levels at grades A-C. No GCSEs. 17. CHRTID/SERIAL: 10418066 BCS 2000 Qualifications: 9 GCSEs, all at A-C grades, 8 obtained in 1986 and 1 in 1987. No O levels or CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 9 O levels, all at A-C grades, no CSEs or GCSEs. 18. CHRTID/SERIAL: 15428033 BCS 2000 Qualifications: 2 GCSEs, both at below grade C, in 1986. No O levels. 3 CSEs in 1986, 2 at grade 1, 1 at lower grade. Present in BCS 1996: No 19. CHRTID/SERIAL: 4169024 BCS 2000 Qualifications: 3 GCSEs in 1987, all at A-C grades; 2 O levels in 1987, at A-C grades, no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 4 O levels, all at below grade C. 20. CHRTID/SERIAL: 4251023 BCS 2000 Qualifications: 1 GCSE in 1987, at A-C grade; 3 O levels in 1986, 2 at A-C grades, 1 at lower grade; 6 CSEs in 1986, 2 at grade 1, 4 at lower grades. Present in BCS 1996: No 21. CHRTID/SERIAL: 16306031 BCS 2000 Qualifications: 2 GCSEs in 1986, 1 at A-C grades, 1 at a lower grade. 9 O levels in 1986, 7 at A-C grades, 2 at lower grades. No CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 8 O levels at A-C; coded as –3 for both O levels below grade C and GCSEs.


22. CHRTID/SERIAL: 921022 BCS 2000 Qualifications: 3 GCSEs in 1986, 2 at A-C grades, 1 at lower grade; no O levels; 2 CSEs in 1986, both at below grade 1. Present in BCS 1996: No 23. CHRTID/SERIAL: 16442012 BCS 2000 Qualifications: 7 GCSEs in 1987, 3 at A-C grades, 4 lower grades; 8 O levels in 1987, 3 at A-C grades, 5 at lower grades; 7 CSEs in 1986, 3 at grade 1, 4 at lower grades. Present in BCS 1996: Yes BCS 1996 Qualifications: No CSEs or GCSEs, 2 O levels at A-C grades. 24. CHRTID/SERIAL: 11496003 BCS 2000 Qualifications: 2 GCSEs, 1 in 1986, 1 in 1987, 1 of these is at A-C grade, other at lower grade. No O levels; 4 CSEs in 1986, all at grade 1. Present in BCS 1996: Yes BCS 1996 Qualifications: 3 CSEs at grade 1; 2 CSEs at lower grades; also 1 GCSE at grades A-C. 25. CHRTID/SERIAL: 12954012 BCS 2000 Qualifications: 2 GCSEs in 1986, both at A-C grades. 7 O levels, 2 in 1985, 5 in 1986. 5 of them at A-C grades, 2 at lower grades. No CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 7 O levels at A-C grades, 2 O levels at lower grades; no CSEs. 26. CHRTID/SERIAL: 6649014 BCS 2000 Qualifications: 11 GCSEs, all at A-C grade, 9 in 1986, 2 in 1987. No O levels, no CSEs. Present in BCS 1996: No 27. CHRTID/SERIAL: 10911025 BCS 2000 Qualifications: 10 GCSEs in 1986, all at A-C grades. No O level, no CSE. Present in BCS 1996: Yes BCS 1996 Qualifications: 9 O levels at A-C grade. 1 GCSE at A-C. 28. CHRTID/SERIAL: 13100091 BCS 2000 Qualifications: 5 GCSEs in 1986. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 5 O levels at A-C grades; no CSEs; no GCSEs. 29. CHRTID/SERIAL: 14836090 BCS 2000 Qualifications: 4 GCSEs in 1986, all at A-C grades; 1 O level in 1986 at A-C grade; 1 CSE in 1986 at grade 1. Present in BCS 1996: No 30. CHRTID/SERIAL: 3845033 BCS 2000 Qualifications: 8 GCSEs in 1986, all at below grade C. No O levels. 3 CSEs in 1986, at below grade 1. Present in BCS 1996: Yes BCS 1996 Qualifications: Has ticked the no qualifications box. 31. CHRTID/SERIAL: 16997056 BCS 2000 Qualifications: 11 GCSEs in 1986, 10 at A-C grades, 1 at a lower grade. No O levels, no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: Also 10 GCSEs at A-C, and 1 at a lower grade. No O levels or CSEs. 32. CHRTID/SERIAL: 12429074 BCS 2000 Qualifications: 7 GCSEs in 1986, 6 at A-C grades, 1 at a lower grade. No O levels or CSEs. Present in BCS 1996: Yes


BCS 1996 Qualifications: 2 CSEs at grade 1; 6 CSEs at other grades, 1 O level at A-C. No GCSEs. 33. CHRTID/SERIAL: 15685019 BCS 2000 Qualifications: 3 GCSEs in 1986, all at A-C grades. No O levels or CSEs. Present in BCS 1996: No 34. CHRTID/SERIAL: 14449083 BCS 2000 Qualifications: 1 GCSE at A-C grade in 1986. No O levels or CSEs. Present in BCS 1996: No 35. CHRTID/SERIAL: 16441037 BCS 2000 Qualifications: 8 GCSEs in 1986, 6 at A-C grades, 2 at lower grades. No O levels or CSEs. Present in BCS 1996: No 36. CHRTID/SERIAL: 6990050 BCS 2000 Qualifications: 8 GCSEs in 1986, 6 at A-C grades, 2 at lower grades. 1 O level at A-C grade in 1987. No CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 8 CSEs all below grade 1; 5 O levels at A-C grades; no GCSEs. 37. CHRTID/SERIAL: 10291095 BCS 2000 Qualifications: 8 GCSEs, 7 of them obtained in 1986 and 1 in 1987, all of them were at A-C grades. 8 O levels, 7 in 1986 and 1 in 1987, all at A-C grades. No CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 8 O levels at A-C grades. 38. CHRTID/SERIAL: 12162020 BCS 2000 Qualifications: 7 GCSEs, 6 in 1986, and 1 in 1988, 4 of them at A-C grades and 3 at lower grades. No O levels or CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 3 O levels at A-C grades, 1 further O level at a lower grade; 5 CSEs all below grade 1. No GCSEs. 39. CHRTID/SERIAL: 20344000 BCS 2000 Qualifications: 6 GCSEs in 1986, 4 at A-C grades, 2 at lower grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 1 CSE grade 1, and 4 CSEs at lower grades; 1 O level at below grade C. No GCSEs. 40. CHRTID/SERIAL: 2796034 BCS 2000 Qualifications: 5 GCSEs, 3 in 1986 and 2 in 1988, all at A-C grades. 1 O level in 1987, at A-C grade. 6 CSEs in 1986, 1 at grade 1 and 5 at lower grades. Present in BCS 1996: No 41. CHRTID/SERIAL: 11994039 BCS 2000 Qualifications: 8 GCSEs in 1986, 4 at A-C grades, and 4 at lower grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 8 CSEs, all at below grade 1; No O levels and no GCSEs. 42. CHRTID/SERIAL: 12873012 BCS 2000 Qualifications: 4 GCSEs in 1987, 3 at A-C and 1 at a lower grade. 2 O levels in 1988 both at A-C grades, no CSEs. Present in BCS 1996: No


43. CHRTID/SERIAL: 14521254 BCS 2000 Qualifications: 5 GCSEs in 1986, all at below grade C; 1 O level in 1986 at A-C grade, no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 3 CSEs at grade 1; no O levels or GCSEs. 44. CHRTID/SERIAL: 5627033 BCS 2000 Qualifications: 3 GCSEs, in 1986, 2 at A-C grades and 1 at a lower grade. 5 O levels in 1986, all at A-C grades; 2 CSEs in 1986, 1 of them at grade 1 and 1 at a lower grade. Present in BCS 1996: Yes BCS 1996 Qualifications: 5 O levels at A-C; 3 GCSEs at A-C 1 of them at A-C and 1 at a lower grade; 2 CSEs – 1 at grade 1, 1 at a lower grade. 45. CHRTID/SERIAL: 17770000 BCS 2000 Qualifications: 6 GCSEs in 1986, 5 at A-C grades and 1 at a lower grade. 1 O level in 1986 at below C grade; no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 5 CSEs at below grade 1, 1 O level below grade C; no GCSEs. 46. CHRTID/SERIAL: 6897000 BCS 2000 Qualifications: 8 GCSEs in 1986, 6 at A-C grades, 2 at lower grades; 8 O levels in 1986, 6 at A-C, 2 at lower grades. No CSEs. Present in BCS 1996: No 47. CHRTID/SERIAL: 120029 BCS 2000 Qualifications: 2 GCSEs in 1986, both at A-C grades; 8 O levels, 6 in 1986, 2 in 1987, all at A-C grades. No CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 11 O levels at A-C grades; 1 GCSE at A-C grade. 48. CHRTID/SERIAL: 2036036 BCS 2000 Qualifications: 8 GCSEs in 1986, 7 at A-C grades, 1 at a lower grade. No O levels, 1 CSE at grade 1. Present in BCS 1996: Yes BCS 1996 Qualifications: 1 CSE grade 1, 4 O levels at A-C, 2 GCSEs at A-C grade. 49. CHRTID/SERIAL: 5043021 BCS 2000 Qualifications: . 9 GCSEs, with 8 in 1986 and 1 in 1991, all at A-C grades. 1 O level in 1990 at A-C grade; no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 9 O levels at A-C grades, no GCSEs or CSEs. 50. CHRTID/SERIAL: 12989017 BCS 2000 Qualifications: 1 GCSE in 1987, at A-C grade; 8 O levels in 1986, 7 at A-C grades, 1 at a lower grade; 1 CSE in 1986 at grade 2 or lower. Present in BCS 1996: Yes BCS 1996 Qualifications: 1 CSE at grade 2 or below; 7 O levels at A-C grades, 1 O level at a lower grade; 1 GCSE at A-C grade. 51. CHRTID/SERIAL: 11458049 BCS 2000 Qualifications: 4 GCSEs, all at A-C grades, 2 obtained in 1986 and 2 in 1987. No O levels. 4 CSEs in 1986, all at below grade 1. Present in BCS 1996: Yes BCS 1996 Qualifications: 4 GCSEs at A-C grades; 5 GCSEs at other grades. Note left school and full time education at 17, so unlikely to have GCSEs.


52. CHRTID/SERIAL: 14811061 BCS 2000 Qualifications: 1 GCSE in 1987, at A-C grade. 10 O levels in 1985, 9 at A-C grades, 1 at a lower grade. No CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 9 O levels all at A-C grades, no GCSEs. 53. CHRTID/SERIAL: 15304002 BCS 2000 Qualifications: 3 GCSEs in 1987, all at A-C grades; 7 O levels in 1986, all at A-C grades; 4 CSEs in 1986, all at grade 1. Present in BCS 1996: Yes BCS 1996 Qualifications: 2 CSEs at grade 1; 8 O levels at A-C grades; 2 GCSEs at A-C grades. 54. CHRTID/SERIAL: 13760011 BCS 2000 Qualifications: 5 GCSEs, all at A-C grades, 2 obtained in 1987 and 3 in 1986. No O levels or CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 7 CSEs – all at below grade 1; 5 O levels at A-C grades; no GCSEs. 55. CHRTID/SERIAL: 13970042 BCS 2000 Qualifications: 4 GCSEs in 1986, all at A-C grades; no O levels or CSEs. Present in BCS 1996: No 56. CHRTID/SERIAL: 13624030 BCS 2000 Qualifications: 7 GCSEs in 1987, 2 at A-C grades and 5 at lower grades. Present in BCS 1996: No 57. CHRTID/SERIAL: 11025086 BCS 2000 Qualifications: 3 GCSEs, 2 in 1987 and 1 in 1986, 2 at A-C grades, 1 at a lower grade. Present in BCS 1996: Yes BCS 1996 Qualifications: 1 CSE at below grade 1; 4 O levels at A-C grades; 2 GCSEs at A-C grades. 58. CHRTID/SERIAL: 10086040 BCS 2000 Qualifications: 8 GCSEs, all at A-C grades, 1 in 1985 and 7 in 1986. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 8 O levels at A-C grades; no CSEs or GCSEs. 59. CHRTID/SERIAL: 5048098 BCS 2000 Qualifications: 9 GCSEs in 1986, all at A-C grades. No CSEs or O levels. Present in BCS 1996: Yes BCS 1996 Qualifications: 9 O levels at A-C grades, no GCSEs. 60. CHRTID/SERIAL: 12324020 BCS 2000 Qualifications: 8 GCSEs (5 at A-C grade, 3 at lower grades), 5 of them in 1986, 2 in 1988 and 1 in 1993. No O levels or CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 4 CSEs: 1 at grade 1, 3 at lower grades; 1 O level at A-C grade; 3 GCSEs at A-C grades. 61. CHRTID/SERIAL: 16488069 BCS 2000 Qualifications: 7 GCSEs in 1986, 4 at A-C grades, 3 at lower grades. No O levels and no CSEs. Present in BCS 1996: No 62. CHRTID/SERIAL: 1270066 BCS 2000 Qualifications: 5 GCSEs in 1985, all at A-C grades; 2 CSEs in 1986, both below grade 1; no O levels.


Present in BCS 1996: Yes BCS 1996 Qualifications: 5 O levels at A-C grades; no GCSEs or CSEs. 63. CHRTID/SERIAL: 10165090 BCS 2000 Qualifications: 10 GCSEs in 1986, all at A-C grades; no O levels or CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 10 O levels at A-C grades, no CSEs or GCSEs. 64. CHRTID/SERIAL: 10263265 BCS 2000 Qualifications: 9 GCSEs (6 at A-C grades, 3 at lower grades), 8 obtained in 1986 and 1 in 1990. Present in BCS 1996: Yes BCS 1996 Qualifications: 2 CSEs: 1 at grade 1, 1 at a lower grade; 4 O levels at A-C grades, 3 O levels at lower grades; 1 GCSE at A-C grade. 65. CHRTID/SERIAL: 2619073 BCS 2000 Qualifications: 7 GCSEs, all at A-C grades, 5 in 1986 and 2 in 1987. 1 O level, A-C grade, in 1985; 1 CSE in 1986 at below grade 1. Present in BCS 1996: No 66. CHRTID/SERIAL: 11731001 BCS 2000 Qualifications: 9 GCSEs in 1986, 3 at A-C grades, 6 at lower grades. Present in BCS 1996: Yes BCS 1996 Qualifications: 3 GCSEs at A-C grades and 6 at lower grades. Note age of leaving school and full-time education is recorded as 16. 67. CHRTID/SERIAL: 1552096 BCS 2000 Qualifications: 9 GCSEs in 1986, all at A-C grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: No information on qualifications. 68. CHRTID/SERIAL: 3678033 BCS 2000 Qualifications: 3 GCSEs: 1 in 1986, 1 in 1987, 1 in 1989, all at A-C grades. 8 O levels, all A-C grades, 7 obtained in 1986 and 1 in 1987. No CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 10 O levels, all at below grade C; 1 GCSE at A-C grade. 69. CHRTID/SERIAL: 11878010 BCS 2000 Qualifications: 8 GCSEs (5 at A-C grades, 3 at lower grades): 4 in 1986, 2 in 1988, 1 in 1989, 1 in 1994. No O levels or CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: No information on qualifications. 70. CHRTID/SERIAL: 6807013 BCS 2000 Qualifications: 4 GCSEs in 1986, 3 at A-C grades, 1 at a lower grade. No O levels or CSEs. Present in BCS 1996: No 71. CHRTID/SERIAL: 8027004 BCS 2000 Qualifications: 4 GCSEs (3 at A-C grades, 1 lower grade), 1 in 1986, 1 in 1987, 2 in 1989; 6 O levels in 1986, 1 at A-C grade and 5 at lower grades; no CSEs. Present in BCS 1996: No 72. CHRTID/SERIAL: 2137089 BCS 2000 Qualifications: 2 GCSEs, both at A-C grades, 1 in 1987 and 1 in 1988. 3 O levels (2 at A-C grades and 1 at lower grade), 2 in 1986 and 1 in 1987; 6 CSEs in 1986, all below grade 1. Present in BCS 1996: Yes


BCS 1996 Qualifications: 6 CSEs, all below grade 1; 1 O level at A-C grade; 3 GCSEs at A-C grade. 73. CHRTID/SERIAL: 4197002 BCS 2000 Qualifications: 1 GCSE in 1986 at A-C grade; 6 O levels in 1986, 5 at A-C grades, and 1 at a lower grade; no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 7 O levels: 6 at A-C grades and 1 at a lower grade; 1 CSE at below grade 1. 74. CHRTID/SERIAL: 2772081 BCS 2000 Qualifications: 1 GCSE in 1986, at A-C grade. No O levels or CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-time education at 16; -3 for CSEs at below grade 1 ??. No other qualifications recorded. 75. CHRTID/SERIAL: 12690085 BCS 2000 Qualifications: 1 GCSE in 1986, at A-C grades; 4 O levels in 1996, 3 at A-C grades and 1 at a lower grade; 3 CSEs in 1986, 1 at grade 1, 2 at lower grades. Present in BCS 1996: Yes BCS 1996 Qualifications: 2 CSEs: 1 at grade 1, 1 at a lower grade. 5 O levels: 4 of them at A-C grades, 1 at a lower grade. No GCSEs. 76. CHRTID/SERIAL:15338032 BCS 2000 Qualifications: 7 GCSEs in 1986, all at A-C grades; 1 O level in 1986, at A-C grade; no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 7 O levels at A-C grades; 1 CSE grade 1, no GCSEs. 77. CHRTID/SERIAL: 1714096 BCS 2000 Qualifications: 9 GCSEs, 8 in 1986, 1 in 1987; no O levels and no CSEs. Summary data says 10 GCSEs, all at A-C grades. Present in BCS 1996: No 78. CHRTID/SERIAL: 9419066 BCS 2000 Qualifications: 6 GCSEs in 1986, 5 at A-C grades, 1 at lower grade. No O levels and no CSEs. Present in BCS 1996: No 79. CHRTID/SERIAL: 14358006 BCS 2000 Qualifications: 8 GCSEs (6 at A-C grades, 2 at lower grades), 7 in 1986, 1 in 1987. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 6 GCSEs at A-C grades. 80. CHRTID/SERIAL: 5585063 BCS 2000 Qualifications: 3 GCSEs in 1986, all at A-C grades; no O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: (Left school and full-time education at 16, making it very unlikely that any GCSEs). 1 CSE at below grade 1. 81. CHRTID/SERIAL: 9222061 BCS 2000 Qualifications: 2 GCSEs in 1986, both at A-C grades. 2 O levels in 1986 at A-C grades; 1 CSE in 1986 at grade 1. Present in BCS 1996: Yes BCS 1996 Qualifications: (left school and full-time education in 1986). 3 O levels at A-C grades; 3 GCSEs at A-C grades.


82. CHRTID/SERIAL: 14487037 BCS 2000 Qualifications: 2 GCSEs in 1986, both at A-C grades; no O levels or CSEs. Present in BCS 1996: No 83. CHRTID/SERIAL: 1504065 BCS 2000 Qualifications: 9 GCSEs in 1986, all at A-C grades. No O levels or CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: No information on qualifications 84. CHRTID/SERIAL: 5348029 BCS 2000 Qualifications: 9 GCSEs, 8 in 1986 and 1 in 1994. Summary says 7 GCSEs, 7 at A-C and 2 at lower grades. No O levels or CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 1 CSE at below grade 1, left school and full-time education at 16. 85. CHRTID/SERIAL: 5793043 BCS 2000 Qualifications: 3 GCSEs in 1985, all at A-C grades. No O levels and no CSEs. Present in BCS 1996: No 86. CHRTID/SERIAL: 9219011 BCS 2000 Qualifications: 9 GCSEs, (7 at A-C grades, 2 at lower grades), 7 in 1986, 2 in 1987; no O levels or CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: No information on qualifications. 87. CHRTID/SERIAL: 1367094 BCS 2000 Qualifications: 5 GCSES in 1986, 4 at A-C and 1 at lower grade. No O levels and no CSEs. Present in BCS 1996: No 88. CHRTID/SERIAL: 12163096 BCS 2000 Qualifications: 10 GCSEs, all at A-C grades, 8 in 1986, 1 in 1987, 1 in 1988. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: No information on qualifications 89. CHRTID/SERIAL: 14486062 BCS 2000 Qualifications: 6 GCSEs in 1986, 3 at A-C grades, and 3 at lower grades. 2 O levels in 1986, both at A-C grades; no CSEs. Present in BCS 1996: No 90. CHRTID/SERIAL: 15934095 BCS 2000 Qualifications: 2 GCSEs in 1986, 1 at A-C grades and 1 at lower grades. Present in BCS 1996: Yes BCS 1996 Qualifications: 2 O levels at A-C; no GCSEs. 91. CHRTID/SERIAL: 3935034 BCS 2000 Qualifications: 5 GCSEs, all at A-C grades, 4 in 1987 and 1 in 1989. 7 O levels in 1986, all at A-C grades; no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: No information on qualifications 92. CHRTID/SERIAL: 15648040 BCS 2000 Qualifications: 7 GCSEs, all at A-C grades; 6 in 1986 and 1 in 1991. No O levels and no CSEs. Present in BCS 1996: No


93. CHRTID/SERIAL: 10366019 BCS 2000 Qualifications: 6 GCSEs in 1986, all at A-C grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: No information on qualifications. 94. CHRTID/SERIAL: 9232037 BCS 2000 Qualifications: 8 GCSEs in 1986, all at A-C grades. Present in BCS 1996: No 95. CHRTID/SERIAL: 11337044 BCS 2000 Qualifications: 5 GCSEs in 1986, all at A-C grades. No O levels and no CSEs Present in BCS 1996: Yes BCS 1996 Qualifications: No information on qualifications. (Left school and full-time education at 17, before introduction of GCSEs). 96. CHRTID/SERIAL: 1412013 BCS 2000 Qualifications: 7 GCSEs: 5 in 1986 and 2 in 1987. Summary data states 8 GCSEs, 6 at A-C grades and 2 lower grades. No O levels and no CSEs Present in BCS 1996: No 97. CHRTID/SERIAL: 9519043 BCS 2000 Qualifications: 1 GCSE in 1986, at A-C grade. No O levels and no CSEs Present in BCS 1996: Yes BCS 1996 Qualifications: 4 GCSEs at below grade C. 98. CHRTID/SERIAL:13705030 BCS 2000 Qualifications: 4 GCSEs in 1984, 1 at A-C grade, 3 at lower grades. No O levels and no CSEs Present in BCS 1996: Yes BCS 1996 Qualifications: No information on qualifications. Left school and full-time education at 16. 99. CHRTID/SERIAL: 9064062 BCS 2000 Qualifications: 7 GCSEs (6 at A-C grades, 1 at lower grade), 6 in 1986 and 1 in 1987. No O levels and no CSEs Present in BCS 1996: Yes BCS 1996 Qualifications: No information on qualifications. 100. CHRTID/SERIAL: 1701094 BCS 2000 Qualifications: 1 GCSE in 1986 at A-C grade; 4 O levels in 1986, 2 at A-C grades and 2 at lower grades. No CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: No information on qualifications. 101. CHRTID/SERIAL: 12625027 BCS 2000 Qualifications: 1 GCSE in 1986, at A-C grade. 3 O levels in 1986, all at A-C grades; no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: No information on qualifications. 102. CHRTID/SERIAL: 10168015 BCS 2000 Qualifications: 1 GCSE in 1986, at A-C grades; 3 O levels in 1986, all at A-C grades; 4 CSEs in 1986, all at grade 1. Present in BCS 1996: Yes BCS 1996 Qualifications: 4 CSEs at grade 1; 4 CSEs at other grades; 3 O levels at A-C grades; 3 GCSEs, 2 at A-C grades, 1 at a lower grade. 103. CHRTID/SERIAL: 1674076


BCS 2000 Qualifications: 7 GCSEs in 1986, 4 at A-C grades, 3 at lower grades. No O levels and no CSEs. Present in BCS 1996: No 104. CHRTID/SERIAL: 14718010 BCS 2000 Qualifications: 6 GCSEs in 1986, all at A-C grades. No O levels and no CSEs. Present in BCS 1996: No 105. CHRTID/SERIAL: 13305021 BCS 2000 Qualifications: 1 GCSE in 1987, at A-C grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 5 CSEs, 2 at grade 1, 3 at other grades; 7 O levels, 5 at A-C grades, 2 at other grades. 1 GCSE at A-C grade. 106. CHRTID/SERIAL: 13764012 BCS 2000 Qualifications: 4 GCSEs in 1986, all at A-C grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 3 CSEs, 1 at grade 1, 2 at other grades; 5 O levels, 2 at A-C and 3 at other grades. No GCSEs. 107. CHRTID/SERIAL: 16291064 BCS 2000 Qualifications: 2 GCSEs in 1987, both at A-C grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 2 GCSEs at A-C grades, no other qualifications. 108. CHRTID/SERIAL: 17341038 BCS 2000 Qualifications: 7 GCSEs in 1986, 3 at A-C grades; 4 at lower grades. No O levels and no CSEs. Present in BCS 1996: No 109. CHRTID/SERIAL: 17437091 BCS 2000 Qualifications: 4 GCSEs in 9186, all at A-C grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 4 CSEs: 1 at grade 1, 3 at other grades. No GCSEs or O levels. 110. CHRTID/SERIAL: 12605075 BCS 2000 Qualifications: 1 GCSE in 1986, at A-C grade; 4 O levels in 1986, 1 at A-C grade, 3 at lower grades; 4 CSEs in 1986, all below grade 1. Present in BCS 1996: Yes BCS 1996 Qualifications: Left school and full-time education at 16; 5 CSEs all below grade 1; 3 O levels, 2 at A-C grades, 1 at lower grade. No GCSEs. 111. CHRTID/SERIAL: 5905062 BCS 2000 Qualifications: 2 GCSEs in 1986, both at A-C grades. No O levels, 4 CSEs in 1986, all at grade 1. Present in BCS 1996: No 112. CHRTID/SERIAL: 11919005 BCS 2000 Qualifications: 10 GCSEs in 1986, 9 at A-C grades, 1 at a lower grade. 10 O levels in 1986, 9 at A-C and 1 at lower grade. No CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 8 O levels at A-C grade, 1 CSE grade 1. No GCSEs. 113. CHRTID/SERIAL: 13335050 BCS 2000 Qualifications: 6 GCSEs in 1986, 4 at A-C and 2 at lower grades. No O levels and no CSEs.


Present in BCS 1996: Yes BCS 1996 Qualifications: Left school and full-time education at age 16; 5 GCSEs, 2 at A-C grade; 3 at lower grades; no O levels or CSEs. 114. CHRTID/SERIAL: 11676005 BCS 2000 Qualifications: 2 GCSEs in 1986, 1 at A-C and 1 at lower grade. 2 O levels in 1986, 1 at A-C and 1 at lower grade; 1 CSE in 1986 at grade 1. Present in BCS 1996: Yes BCS 1996 Qualifications: 1 CSE grade 1; 3 O levels at A-C grades. No GCSEs. 115. CHRTID/SERIAL: 8604090 BCS 2000 Qualifications: 7 GCSEs in 1986, 5 at A-C grades, and 2 at lower grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-timed education at 16, 3 CSEs at grade 1, 4 O levels, all below grade C. 116. CHRTID/SERIAL: 16639266 BCS 2000 Qualifications: 8 GCSEs in 1986, 2 at A-C grades, and 6 at lower grades. No O levels and no CSEs. Present in BCS 1996: No 117. CHRTID/SERIAL: 14676040 BCS 2000 Qualifications: 1 GCSE in 1987, at A-C grade. 7 O levels in 1986, all at A-C grade; no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 8 O levels at A-C grades; 1 CSE at below grade 1. 118. CHRTID/SERIAL: 15215077 BCS 2000 Qualifications: 6 GCSEs in 1986, all at A-C grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: (left school and full-time education at age 16), 6 O levels at A-C grades, no GCSEs. 119. CHRTID/SERIAL: 262062 BCS 2000 Qualifications: 4 GCSEs in 1986, all at A-C grades. Present in BCS 1996: Yes BCS 1996 Qualifications: claims to have left school and full-time education at age 15, and has ticked the box for ‘no qualifications’. 120. CHRTID/SERIAL: 12693010 BCS 2000 Qualifications: 1 GCSE in 1986, at A-C grade; 7 O levels in 1986, with 6 at A-C grades and 1 at a lower grade. No CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 6 O levels at A-C grades, 1 GCSE at A-C grade. 121. CHRTID/SERIAL: 10441068 BCS 2000 Qualifications: 4 GCSEs in 1986, all at A-C grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-time education at age 16, 5 CSEs all at below grade 1, no GCSEs. 122. CHRTID/SERIAL: 15814065 BCS 2000 Qualifications: 6 GCSEs in 1986, 3 at A-C grades and 3 at lower grades. No O levels and no CSEs. Present in BCS 1996: No 123. CHRTID/SERIAL: 16007075


BCS 2000 Qualifications: 2 GCSEs in 1986, both at A-C grades. No O levels, 1 CSE in 1986, at below grade 1. Present in BCS 1996: Yes BCS 1996 Qualifications: 3 CSEs at below grade 1, no GCSEs. 124. CHRTID/SERIAL: 5773091 BCS 2000 Qualifications: 3 GCSEs in 1986, all at A-C grades. No O levels; 4 CSEs in 1986, 1 at grade 1, 3 at lower grades. Present in BCS 1996: Yes BCS 1996 Qualifications: 7 GCSEs at below grade C; no O levels or CSEs. 125. CHRTID/SERIAL: 10481073 BCS 2000 Qualifications: 6 GCSEs in 1986; summary data suggests 9 GCSEs, 8 at A-C grades, 1 at a lower grade. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 8 O levels at A-C grades; no GCSEs. 126. CHRTID/SERIAL: 2832053 BCS 2000 Qualifications: 6 GCSEs in 1986; summary data suggests 9 GCSEs all at A-C grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 9 O levels at A-C grades; no GCSEs. 127. CHRTID/SERIAL: 14225075 BCS 2000 Qualifications: 1 GCSE in 1987, at below grade C. No O levels. 4 CSEs in 1986, all at grade 1. Present in BCS 1996: Yes BCS 1996 Qualifications: 3 CSEs, 2 at grade 1, and 1 at a lower grade; 1 O level at A-C grade; 1 GCSE at below grade C. 128. CHRTID/SERIAL: 2816001 BCS 2000 Qualifications: 3 GCSEs in 1987, all at A-C grades. No O levels and no CSEs. Present in BCS 1996: No 129. CHRTID/SERIAL: 7103050 BCS 2000 Qualifications: 8 GCSEs in 1986, 4 at A-C grades and 4 at lower grades. No O levels and no CSEs. Present in BCS 1996: No 130. CHRTID/SERIAL: 3522023 BCS 2000 Qualifications: 1 GCSE in 1986 at below grade C; 6 O levels in 1986, 2 at A-C grades and 4 at lower grades. 7 CSEs in 1986, 2 at grade 1 and 5 at lower grades. Present in BCS 1996: Yes BCS 1996 Qualifications: 7 CSEs, 2 at grade 1 and 5 at lower grades; 6 O levels, 2 at A-C grades and 4 at lower grades. No GCSEs. 131. CHRTID/SERIAL: 3524074 BCS 2000 Qualifications: 4 GCSEs in 1986, 3 at A-C grades and 1 at a lower grade. No O levels and no CSEs. Present in BCS 1996: No 132. CHRTID/SERIAL: 3323044 BCS 2000 Qualifications: 8 GCSEs in 1986, 1 at A-C grades and 7 at lower grades. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-time education at age 16; 7 CSEs all below grade 1; no O levels or GCSEs.


133. CHRTID/SERIAL: 160034 BCS 2000 Qualifications: 5 GCSEs in 1986, 5 at A-C grades and 1 at a lower grade. 2 O levels in 1986, both at A-C grades; no CSEs. Present in BCS 1996: No 134. CHRTID/SERIAL: 2740001 BCS 2000 Qualifications: 6 GCSEs in 1986, all at A-C grades. 1 O level in 1986, at A-C grades, no CSEs. Present in BCS 1996: No 135. CHRTID/SERIAL: 15240029 BCS 2000 Qualifications: 8 GCSEs, 4 in 1986 and 4 in 1987, all at A-C grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 10 O levels, 8 at A-C grades and 2 at lower grades; 4 CSEs, 3 at grade 1 and 1 at a lower grade; no GCSEs. 136. CHRTID/SERIAL: 3463026 BCS 2000 Qualifications: 3 GCSEs in 1986, 2 at A-C grades and 1 at a lower grade. 5 O levels in 1986, all at A-C grades; 4 CSEs in 1986, 3 at grade 1 and 1 at a lower grade. Present in BCS 1996: Yes BCS 1996 Qualifications: 5 O levels at A-C grade; 4 CSEs, 3 at grade 1 and 1 at a lower grade; 3 GCSEs, 2 at A-C grades and 1 at a lower grade. 137. CHRTID/SERIAL: 8722069 BCS 2000 Qualifications: 8 GCSEs in 1986, all at A-C grades. No O levels and no CSEs. Present in BCS 1996: No 138. CHRTID/SERIAL: 59058 BCS 2000 Qualifications: 6 GCSEs (2 at A-C grade and 4 at lower grades), 5 in 1986 and 1 in 1987. Present in BCS 1996: Yes BCS 1996 Qualifications: 7 CSEs, 2 at grade 1 and 5 at other grades; 6 O levels, 3 at A-C grades and 3 at other grades; 1 GCSE at A-C grade. 139. CHRTID/SERIAL: 6565089 BCS 2000 Qualifications: 2 GCSEs in 1986, both at A-C grades; no O levels; 3 CSEs in 1986, 1 at grade 1 and 2 at other grades. Present in BCS 1996: No 140. CHRTID/SERIAL: 83035 BCS 2000 Qualifications: 6 GCSEs in 1986, all at A-C grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-time education at 16; 6 CSEs, with 2 at grade 1 and 4 at other grades; no O levels and no GCSEs. 141. CHRTID/SERIAL: 6912091 BCS 2000 Qualifications: 9 GCSEs in 1986; 4 at A-C grades and 5 at other grades; No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 3 CSEs, 1 at grade 1 and 2 at lower grades; 4 O levels at A-C grades; 1 GCSE at A-C grade. 142. CHRTID/SERIAL: 10235038 BCS 2000 Qualifications: 6 GCSEs in 1986, 5 at A-C grades and 1 at a lower grade. No O levels; 5 CSEs, all at grade 1, 2 in 1991 and 3 in 1992. Present in BCS 1996: Yes


BCS 1996 Qualifications: 8 CSEs, 1 at grade 1; 7 at lower grades; 5 GCSEs, 3 at A-C grades and 2 at lower grades. 143. CHRTID/SERIAL: 11719051 BCS 2000 Qualifications: 6 GCSEs, 3 in 1986, 1 in 1992, 2 in 1993; summary data suggests 9 GCSEs, 2 at A-C grades and 7 at lower grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 6 CSEs, all below grade 1; 2 O levels at A-C grades; 4 GCSEs, 2 at A-C grades and 2 at lower grades. 144. CHRTID/SERIAL: 9515042 BCS 2000 Qualifications: 6 GCSEs in 1986, all at A-C grades. Present in BCS 1996: Yes BCS 1996 Qualifications: 1 CSE at grade 1; 5 O levels at A-C grades; 4 GCSEs at below grade C. 145. CHRTID/SERIAL: 10695004 BCS 2000 Qualifications: 1 GCSE in 1986, at below grade C; 6 O levels in 1986, 4 at A-C grades, 2 at lower grades; 2 CSEs in 1986, 1 at grade 1, and 1 at other grade. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-time education at age 16; 2 CSEs, 1 at grade and 1 at lower grade; 5 O levels, 4 at A-C grade, 1 at other grade; 1 GCSE at below grade C. 146. CHRTID/SERIAL: 8599075 BCS 2000 Qualifications: 4 GCSEs in 1987, all at A-C grades; 2 O levels in 1986, 1 at A-C grades, and 1 at lower grade; 5 CSEs in 1986, 2 at grade 1 and 3 at lower grades. Present in BCS 1996: No 147. CHRTID/SERIAL: 9560047 BCS 2000 Qualifications: 9 GCSEs, 7 in 1986 and 2 in 1987; 6 at A-C grades and 3 at lower grades. No O levels and no CSEs. Present in BCS 1996: No 148. CHRTID/SERIAL: 2988063 BCS 2000 Qualifications: 7 GCSEs in 1986, all at below grade C. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 7 CSEs, all at below grade 1. No O levels and no GCSEs. 149. CHRTID/SERIAL: 11925081 BCS 2000 Qualifications: 8 GCSEs in 1986, 6 at A-C grades and 2 at other grades. Present in BCS 1996: Yes BCS 1996 Qualifications: 5 O levels at A-C grades, and 2 CSEs at grade 1. No GCSEs. 150. CHRTID/SERIAL: 189064 BCS 2000 Qualifications: 9 GCSEs in 1986, 4 at A-C grades and 5 at lower grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 6 CSEs, 3 at grade 1 and 3 at lower grades; 1 O level at A-C grade; no GCSEs. 151. CHRTID/SERIAL: 11862083 BCS 2000 Qualifications: 8 GCSEs in 1986, 4 at A-C grades and 4 at lower grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-time education at age 16, 5 CSEs, 1 at grade 1 and 4 at lower grades; 5 O levels, 1 at A-C and 4 at lower grades; no GCSEs. 152. CHRTID/SERIAL: 16526038 BCS 2000 Qualifications: 2 GCSEs in 1986, both at A-C grades. No O levels and no CSEs.


Present in BCS 1996: No 153. CHRTID/SERIAL: 6135051 BCS 2000 Qualifications: 9 GCSEs in 1986, 5 at A-C grades and 4 at other grades. Present in BCS 1996: Yes BCS 1996 Qualifications: 9 O levels, 5 at A-C grades, and 4 at other grades. No GCSEs. 154. CHRTID/SERIAL: 12418022 BCS 2000 Qualifications: 9 GCSEs in 1986, 8 at A-C grades and 1 at other grade. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-time education at age 16; 7 CSEs, 1 at grade 1 and 6 at lower grades; 1 GCSE at A-C grade. 155. CHRTID/SERIAL: 2961084 BCS 2000 Qualifications: 5 GCSEs in 1986, all at A-C grades; 7 O levels in 1985, 2 at A-C grades and 5 at lower grades. No CSEs. Present in BCS 1996: No 156. CHRTID/SERIAL: 3563003 BCS 2000 Qualifications: 5 GCSEs in 1986, all at A-C grades; No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-time education at age 16; 8 O levels all at A-C grades; No GCSEs or CSEs. 157. CHRTID/SERIAL: 11633075 BCS 2000 Qualifications: 7 GCSEs (5 at A-C grades and 2 at lower grades), 5 in 1986 and 2 in 1989. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-time education at age 16, 2 CSEs at below grade 1; 7 O levels, 2 at A-C grades and 5 at lower grades. No GCSEs. 158. CHRTID/SERIAL: 12218068 BCS 2000 Qualifications: 9 GCSEs in 1986, 8 at A-C grade and 1 at lower grade. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 8 O levels at A-C grades; no GCSEs or CSEs. 159. CHRTID/SERIAL: 2985037 BCS 2000 Qualifications: 5 GCSEs in 1986, 1 at A-C grade and 4 at lower grades. 2 O levels in 1986, both at A-C grades, 1 CSE in 1986 at grade 1. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-time education at 17; 1 CSE grade 1; 6 O levels, 3 at A-C, 3 at other grades. 160. CHRTID/SERIAL: 2986012 BCS 2000 Qualifications: 5 GCSEs in 1987, all at below C grade; 1 O level in 1987, at A-C grade; 8 CSEs all at below grade 1, 5 in 1986 and 3 in 1987. Present in BCS 1996: Yes BCS 1996 Qualifications: 7 CSEs at below grade 1; 4 O levels, 1 at grade A-C, 3 at below C grade; no GCSEs. 161. CHRTID/SERIAL: 3406095 BCS 2000 Qualifications: 3 GCSEs in 1986, all at A-C grades; 4 CSEs in 1986, all below grade 1; No O levels. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-time education at age 16; 5 CSEs at below grade 1, no GCSEs.


162. CHRTID/SERIAL: 4233021 BCS 2000 Qualifications: 6 GCSEs in 1986, 3 at A-C and 3 at other grades. No O levels and no CSEs. Present in BCS 1996: No 163. CHRTID/SERIAL: 8911072 BCS 2000 Qualifications: 5 GCSEs in 1986, 4 at A-C grades and 1 at other grade; no O levels; 1 CSE in 1986, at below grade 1. Present in BCS 1996: Yes BCS 1996 Qualifications: 3 CSEs at below grade 1; 4 GCSEs at A-C grades. 164. CHRTID/SERIAL: 7465090 BCS 2000 Qualifications: 5 GCSEs in 1986, 1 at A-C grade and 4 at other grades. 5 O levels in 1986, 1 at A-C grade and 4 at lower grades; 3 CSEs in 1986, 1 at grade 1 and 2 at lower grades. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-time education at age 16; no information on qualifications. 165. CHRTID/SERIAL: 8935025 BCS 2000 Qualifications: 4 GCSEs in 1986, 2 at A-C grades and 2 at lower grades; No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 3 CSEs all at below grade 1; 4 O levels, 1 at A-C grade, 3 at lower grades; no GCSEs. 166. CHRTID/SERIAL: 4482082 BCS 2000 Qualifications: 2 GCSEs, both at A-C grade, 1 in 1987 and 1 in 1989. 3 O levels in 1986, all at A-C grades; 5 CSEs in 1986, 2 at grade 1 and 3 at lower grades. Present in BCS 1996: No 167. CHRTID/SERIAL: 9735049 BCS 2000 Qualifications: 2 GCSEs in 1986, both at A-C grades; No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-time education at age 16; No O levels and no CSEs; 2 GCSEs at A-C grades. 168. CHRTID/SERIAL: 15607060 BCS 2000 Qualifications: 5 GCSEs in 1987, 3 at A-C grades and 2 at lower grades; 2 O levels in 1986, both at A-C grades; 1 CSE in 1986 at grade 1. Present in BCS 1996: Yes BCS 1996 Qualifications: 4 CSEs at below grade 1; 4 O levels, 1 at A-C grade and 3 at other grades; 1 GCSE at A-C grade. 169. CHRTID/SERIAL: 10367095 BCS 2000 Qualifications: 3 GCSEs in 1986, all at A-C grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and f/t education at age 16; no information on qualifications. 170. CHRTID/SERIAL: 13324099 BCS 2000 Qualifications: 7 GCSEs in 1986, 4 at A-C grades and 3 at lower grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-time education at age 16; 2 CSEs at below grade 1; 4 O levels at A-C grades; 1 GCSE at below grade C. 171. CHRTID/SERIAL: 10214010 BCS 2000 Qualifications: 8 GCSEs in 1986, all at A-C grades. No O levels and no CSEs. Present in BCS 1996: Yes


BCS 1996 Qualifications: 8 O levels at A-C grades; no CSEs and no GCSEs. 172. CHRTID/SERIAL: 6443008 BCS 2000 Qualifications: 11 GCSEs (5 at A-C grades and 6 at other grades), 7 in 1986, 2 in 1987, 2 in 1991. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-time education at age 16; 7 CSEs, 1 at grade 1, and 6 at lower grades; 2 O levels at A-C; 1 GCSE at A-C grade. 173. CHRTID/SERIAL: 10583000 BCS 2000 Qualifications: 7 GCSEs in 1986, 4 at A-C grades and 3 at lower grades; no O levels; 3 CSEs in 1986, all at below grade 1. Present in BCS 1996: Yes BCS 1996 Qualifications: 3 CSEs, 1 at grade 1, 2 at lower grades; 3 O levels at A-C grades. No GCSEs. 174. CHRTID/SERIAL: 10287070 BCS 2000 Qualifications: 10 GCSEs in 1986, all at A-C grades; no O levels; 5 CSEs in 1986, 2 at grade 1 and 3 at lower grades. Present in BCS 1996: Yes BCS 1996 Qualifications: claims to have left school and full-time education at 15; no information on qualifications. 175. CHRTID/SERIAL: 9721072 BCS 2000 Qualifications: 6 GCSEs in 1986, 4 at A-C grades and 2 at lower grades. No O levels and no CSEs. Present in BCS 1996: No 176. CHRTID/SERIAL: 9459071 BCS 2000 Qualifications: 3 GCSEs in 1986, all at A-C grades.No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: (left school and f/t education at 16), 4 CSEs at below grade 1; no O levels or GCSEs. 177. CHRTID/SERIAL: 170010 BCS 2000 Qualifications: 3 GCSEs in 1987, 1 at A-C grade, 2 at lower grades. No O levels and no CSEs. Present in BCS 1996: No 178. CHRTID/SERIAL: 5701006 BCS 2000 Qualifications: 2 GCSEs in 1986, both at A-C grades; no O levels; 4 CSEs in 1986, 1 at grade 1 and 3 at lower grades. Present in BCS 1996: No 179. CHRTID/SERIAL: 174011 BCS 2000 Qualifications: 5 GCSEs in 1986, 3 at A-C grades and 2 at lower grades. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-time education at age 16; 3 CSEs at below grade 1; no O levels or GCSEs. 180. CHRTID/SERIAL: 10778055 BCS 2000 Qualifications: 2 GCSEs in 1987, both at A-C grades. No O levels, 6 CSEs in 1986, all at below grade 1. Present in BCS 1996: No


181. CHRTID/SERIAL: 8561044 BCS 2000 Qualifications: 5 GCSEs in 1986, 2 at A-C grades and 3 at lower grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-time education at age 16. 5 CSEs at below grade 1; 2 O levels at below grade C. No GCSEs. 182. CHRTID/SERIAL: 8471043 BCS 2000 Qualifications: 5 GCSEs in 1986, 4 at A-C grades, 1 at a lower grade. No O levels and no CSEs. Present in BCS 1996: No 183. CHRTID/SERIAL: 5533031 BCS 2000 Qualifications: 5 GCSEs in 1986, 1 at A-C grade, 4 at lower grades; 1 O level in 1986 at A-C grade; 1 CSE in 1986 at grade 1. Present in BCS 1996: No 184. CHRTID/SERIAL: 6072053 BCS 2000 Qualifications: 2 GCSEs, both below grade C, 1 in 1986, 1in 1987. 5 O levels in 1986, 3 at A-C grades, 2 at lower grades; no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 4 CSEs at below grade 1; 6 O levels, 4 at A-C, 2 at other grades; no GCSEs. 185. CHRTID/SERIAL: 12158096 BCS 2000 Qualifications: 1 GCSE in 1986, at A-C grades; 5 O levels in 1986, 2 at A-C grades and 3 at lower grades; no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: No information on qualifications. 186. CHRTID/SERIAL: 8701041 BCS 2000 Qualifications: 10 GCSEs in 1986, 6 at A-C grades and 4 at lower grades. Present in BCS 1996: Yes BCS 1996 Qualifications: 5 CSEs, 1 at grade 1 and 4 at other grades; 3 O levels at A-C grades; no GCSEs. 187. CHRTID/SERIAL: 4493033 BCS 2000 Qualifications: 5 GCSEs, all at A-C grades, 2 in 1987 and 3 in 1988. 2 O levels in 1986, both at A-C grades; no CSEs. Present in BCS 1996: No 188. CHRTID/SERIAL: 4529028 BCS 2000 Qualifications: 8 GCSEs in 1986, 5 at A-C grades and 3 at lower grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-time education at 16; no information on qualifications. 189. CHRTID/SERIAL: 4532078 BCS 2000 Qualifications: 8 GCSEs in 1986, 3 at A-C and 5 at other grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 4 CSEs, 1 at grade 1 and 3 at other grades; 1 O level at A-C grade; no GCSEs. 190. CHRTID/SERIAL: 13050095 BCS 2000 Qualifications: 4 GCSEs in 1986, all at below C grade; No O levels and no CSEs. Present in BCS 1996: Yes


BCS 1996 Qualifications: 5 O levels at below grade C; GCSE below grade C coded as –3. 191. CHRTID/SERIAL: 1443017 BCS 2000 Qualifications: 3 GCSEs in 1986, all at A-C grades; 4 O levels in 1986, 2 at A-C grades and 2 at other grades. Present in BCS 1996: No 192. CHRTID/SERIAL: 12326071 BCS 2000 Qualifications: 9 GCSEs in 1986, 7 at A-C grades, 2 at other grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 10 O levels, 7 at A-C grades, 3 at lower grades; no GCSEs. 193. CHRTID/SERIAL: 1437042 BCS 2000 Qualifications: 6 GCSEs in 1986, all at A-C grades. No O levels and no CSEs. Present in BCS 1996: No 194. CHRTID/SERIAL: 1639047 BCS 2000 Qualifications: 7 GCSEs in 1986, 1 at A-C, 6 at other grades; No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 8 GCSEs at A-C grades; No O levels and no CSEs. 195. CHRTID/SERIAL: 10678078 BCS 2000 Qualifications: 7 GCSEs in 1986, 5 at A-C and 2 at other grades. No O levels. 2 CSEs in 1986, both at grade 1. Present in BCS 1996: No 196. CHRTID/SERIAL: 2567026 BCS 2000 Qualifications: 6 GCSEs in 1986, 4 at A-C grades and 2 at other grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 6 GCSEs at below grade C, No O levels and no CSEs. 197. CHRTID/SERIAL: 13648084 BCS 2000 Qualifications: 2 GCSEs in 1985, both at A-C grades. No O levels and no CSEs. Present in BCS 1996: No 198. CHRTID/SERIAL: 5609031 BCS 2000 Qualifications: 5 GCSEs in 1986, 3 at A-C and 2 at other grades. No O levels; 3 CSEs in 1986, all at below grade 1. Present in BCS 1996: Yes BCS 1996 Qualifications: left school and full-time education at age 16; 4 CSEs, all below grade 1. No GCSEs. 199. CHRTID/SERIAL: 12041190 BCS 2000 Qualifications: 6 GCSEs in 1986, 4 at A-C grades and 2 at other grades. No O levels and no CSEs. Present in BCS 1996: Yes BCS 1996 Qualifications: 5 CSEs, 2 at grade 1 and 3 at other grades; 3 O levels, 2 at A-C grades and 1 at other grade; 1 GCSE at A-C grade. 200. CHRTID/SERIAL: 509063 BCS 2000 Qualifications: 4 GCSEs in 1986, all at A-C grades; No O levels and no CSEs. Present in BCS 1996: No


Report on GCSE date errors, Appendix 1. (Andrew Jenkins) From: Andrew Jenkins [mailto:[email protected]] Sent: Thursday, May 24, 2001 5:15 PM To: Peter Shepherd (E-mail) Subject: cleaning up ncds Peter, Here is a word file containing the actions which I've taken to produce an NCDS 6 dataset which is clean of erroneous dates for GCSEs. All those reporting pre-1988 GCSEs have been altered in the light of the information contained un NCDS 5 and the Exams file. I've tried to adopt a sensible solution for each case. In the few cases where there is literally no data, we have simply recoded to CSEs. I've done this by going into the dataset and changing the variables that need to be changed. I can e-mail you the altered version of NCDS 6 education data in an spss file, or just the 94 cases which have been changed if that would be useful. regards Andrew Dr Andrew Jenkins Research Officer, Centre for the Economics of Education Mathematical Sciences Group Institute of Education University of London 20 Bedford Way London WC1H OAL email [email protected] Tel 020 7612 6663


APPENDIX 7: Highest Qualification Documentation relating to highest qualification produced by the following is reproduced below:

1. Andrew Jenkins (Institute of Education) and Gerald Makepeace (University of Cardiff)

2. Samantha Parsons, Centre for Longitudinal Studies


([email protected])

1. Andrew Jenkins and Gerald Makepeace The table below shows how educational qualifications obtained by NCDS cohort members have been mapped the into the national qualifications framework.

Level of Qualificat

ion

General (Academic)

Vocationally-related (Applied)

Occupational (Vocational)

5 Higher Degree NVQ level 5 PGCE

4 Degree

HE Diploma

BTEC Higher Certificate/Diploma

HNC/HND

NVQ level 4 Professional degree level

qualifications Nursing/paramedic

Other teacher training qualification

City & Guilds Part 4/Career Ext/Full Tech

RSA Higher Diploma 3 A level

AS levels Scottish Highers

Scottish Cert of 6th Year Studies

Advanced GNVQ BTEC National Diploma

ONC/OND

NVQ level 3 City & Guilds Part

3/Final/Advanced Craft RSA Advanced Diploma

Pitmans level 3 2 GCSE grade A*-C

O levels grade A-C O levels grade D-E

CSE grade 1 Scottish standard

grades 1-3 Scottish lower or ordinary grades

Intermediate GNVQ BTEC First Certificate BTEC First Diploma

NVQ level 2 Apprenticeships

City & Guilds Part 2/Craft/Intermediate

City & Guilds Part 1/Other RSA First Diploma

Pitmans level 2

1 GCSE grade D-G CSEs grades 2-5 Scottish standard

grades 4-5 Other Scottish school

qualification

Foundation GNVQ Other GNVQ

NVQ level 1 Other NVQ

Units towards NVQ RSA Cert/Other Pitmans level 1

Other vocational qualificationsHGV


SPSS syntax ************************* DESCRIPTION OF PROGRAM ************************************. * This program derives academic and qualifications using the 2000 data only. * It defines a set of binary variables showing whether an individual has a particular qualification. * These variables could then be used to devise various typologies for training . * We also derive the highest qualification below. * All variables based on data from the 2000 surveys end in 00. * The interpretation of the results differs across observations. ************************ DEFINITION OF VARIABLES ***********************************************. * There are 2 relevant questions. * The first asks 'Have you obtained any of the qualifications on this card since xx'. * If the respondent has the qualification, the second question asks 'How many did you get?' eg 'How many GCE O-levels did you get ..?' 'How many City and Guilds quals have you obtained ..'. * Where possible, we have defined the variables using the responses to the questions asking how many qualifications of a particular type the respondent has. * Often the respondent says none. We could interpret this as follows. The respondent said 'I have the qualification' when asked the first question but on reflection realised that they did not have it. This interpretation is strengthened by the fact that the information on all types of qualification is collected first. * Suppose the Cohort Member (CM) says I have a GCSE, a GCE and a CSE. * When the interviewer goes through the number of qualifications (one qualfication at a time) the CM realises that the GCSE was GCE and one of the CCEs was a GCE. * The CM may know that he/she has a qualification but does not know how many O-levels, CSEs etc he/she has. The CM may respond 98, 99 but 0 seems unlikely if they have the qualification. * 0 is a genuine response. The feed question for certain qualifications refers to all grades while the 'how many questions' refer to different ranges of grades. Consider somebody who has 8 GCSEs all at grade A. They will say they have a gcse in response to the feeder question, but then answer 8 for the number of grade A-Cs (edgcse1) and 0 to for the number of grade D-Es (edgcse2) . * The code below is easily amended to reflect the programmer's preferences. * Typically the response 'I have none of those qualification' (number equals 0) and the missing values for number of qualifications (8,9,98,99) have to be changed. * We have intepreted 8 and 98 as 'Do not know' and 9 and 99 as 'Not Applicable'. * We have set 8 and 98 equal to 1 (ie denoting possession of the qualification) suggesting that these values reflect an inability to remember the exact number of qualifications obtained. * We have set 9 and 99 equal to 0 suggesting that the respondent has considered the case further and decided that he/she does not have the qualification. * The missing value dummies set 9 for edquals2 and vocqual equal to 0. * Not Applicable is an odd response but it seems very close to saying 'No' I do not have any of the quals. ******************** INTERPRETATION OF VARIABLES ****************************************. * The BCS70 qualifications are collected at ages 26 and 30. The age 16 questions refer to future examinations (ie the survey took place before the exams were taken).


* This program uses the age 30 data (from the 2000 survey). The routing questions ask 'Have you obtained any of the qualifications on this card since April 1986'. Since most school exams were taken in the summer, it is therefore a summary of 'have you ever had this qualification'. It could perhaps be improved by comparing it to the age 26 data but there are differences in the form of the question and the method of collecting the data. * The age 26 survey was a postal survey with a smaller sample than the age 30 sample and there is less detail in the questions asked at 26. * NCDS has regularly collected information on qualifications. The questions in 2000 asked 'whether the qualification has been obtained since the age of 33' for individuals in NCDS V and 'whether the qualification has been obtained since the age of 23' for individuals not in NCDS V. * It is therefore a good summary of 'have you got this qualification since you were x years old'. ******************** EXAMPLE AND AN IMPLICATION OF OUR APPROACH ***************************. * Consider the following code for the variable odeg00=1 if he/she has any other degree level qualifications . * We generate this from the answer to 'how many other degree level qualifications do you have?' (numothdg) using the following code. * recode numothdg (missing=0) (0=0) (98=1)(99=0) (1 thru hi=1) into odeg00 . * variable labels odeg00 "Other degree level quals such as graduate membership of professional institutes". * A crosstabs of numothdg by odeg00 shows that there 99 zeros for numothdeg that give odeg00=0 and 1 value of 98 that gives odeg00=1. * We could have defined the alternative measure, ALTodeg, based on the question 'Do you have any of the following qualifications?' If the individual has 'other degree level qualifications', these are recorded as the value 9 in one of the variables edqtyp14 to edqtyp26. The code for this variable is below. * compute ALTodeg=0 . * if (any(9,edqtyp14,edqtyp15,edqtyp16,edqtyp17,edqtyp18,edqtyp19,edqtyp20,edqtyp21, edqtyp22,edqtyp23,edqtyp24,edqtyp25,edqtyp26) ) ALTodeg=1 . * A crosstabs of ALTodeg by odeg00 shows that there are 99 respondents with ALTodeg not equal odeg00 corresponding to the 99 zeros for numothdeg . This is a large number since there are 963 observations with odeg00=1. * Many people have used the definition similar to ALTodeg rather than the present one. ********************* MISSING VALUES ************************************. * We begin by giving missing values the value 0. We later adjust the data for missing values. * Some analysis below shows that there are few genuinely missing values. The feed questions are 'Have you obtained any of the qualifications on this card? with separate cards for educational and vocational questions. The answers to these questions are contained in edquals2 and vocqual. * (Be careful as there are variables with similar names for the proxy returns.) . * The possible answers to the feed questions are Yes, No, 8, 9, missing. * We treat an answer of Yes or No as saying that we have information on whether the interviewee has one or more of the qualifications. * The same 66 observations have missing values for both the educational and vocational questions. * A further 50 of the educational and 44 of the vocational answers are 8 or 9.


* There are 28 values of 8 or 9 for both the educational and vocational questions. * The variables allmiss, acamiss and vocmiss capture the missing answers. * No respondents have 8 or 9 for the feeder question but subsequently report a qualification. * (Based on edqtyp14 and voctyp12 only). ********************************** PROXY INTERVIEWS *******************************. * We have not picked up the observations (perhaps 26) given for proxy interviews. * Proxy returns occur when someone completed the form on behalf of the CM. * The variable identifying a proxy interview is IntWho in the Household Grid and Housing but we did not extract that information and it is not used below. * A measure (proxy=1) of whether the form is completed by a proxy is computed below . * From the data available to us at present, there are 26 interviews completed by proxy. * Be careful because many of the variables only refer to proxy interview eg agelfted, edquals and vocquals are for proxies, agelfte2, edquals2 and vocqual are for interviews completed by CMs. * edqtyp01 to edqtyp13, voctyp01 to voctyp11 are completed by proxies. * edqtyp14 to edqtyp26, voctyp12 to voctyp22 are completed by CMs . *********************************** GCSE ********************************************. * An apparently large proportion of the BCS70 cohort claim to have done GCSEs. They were one of the last cohorts to do GCE/CSE at school. This program does not resolve the issues raised by this although it seems clear that some respondents have not identified the exams that they took properly. ********************* ADAPTING THE CODE **********************************************. * The code be could be simplified by merging the recode and other commands. For example, the opening code below could be written as . * recode edgcse1 edgcse2 (missing=0) (0=0) (98=1) (99=0) (1 thru hi =1) into gcsehi00 gcselo00 . * variable labels csehi00 "Has CM obtained one or more CSEs at grade 1?" * / cselo00 "Has CM obtained one or more CSEs at grades 2 to 5?". * (Be careful. Some variables use 9 as a missing value but 9 is a legitimate value for other variables.) . * The program is divided into sections because the PC on which it was developed could not handle the whole lot in one go. *********************************************************************************************. ***************************** PROGRAM STARTS *************************************. *********************************************************************************************. **********************************************************************************************. ******************************* LOAD DATA **********************************************. **********************************************************************************************. get file = 'e:\00\ncds-bcs70\qca data\data1.sav' . *ghm at cls get file = 'c:\work\ghmtemp\ncds data\data1.sav' . * pjd at cls get file = 'd:\datawork\qca data analysis\data1.sav' . *************************** GCSE, O-LEVELS, CSE ***********************************. recode edgcse1 (missing=0) (0=0) (98=1) (99=0) (1 thru hi =1) into gcsehi00 . recode edgcse2 (missing=0) (0=0) (98=1) (99=0) (1 thru hi =1) into gcselo00 .


* Make sure 98, 99 are before (1 thru hi) if you want them to be 0 or else they will be coded 1. * In these cases, hi is 13 and 9 respectively. variable labels gcsehi00 "Has CM obtained one or more GCSEs at grades A to C ?". variable labels gcselo00 "Has CM obtained one or more GCSEs at grades D to E ?". recode edolev1 (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into olevhi00 . recode edolev2 (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into olevlo00 . * Hi is 15 and 11. variable labels olevhi00 "Has CM obtained one or more O levels at grades A to C?". variable labels olevlo00 "Has CM obtained one or more O levels at grades D to E?". recode edcse1 (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into csehi00 . recode edcse2 (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into cselo00 . * Hi is 12 and 11. variable labels csehi00 "Has CM obtained one or more CSEs at grade 1?". variable labels cselo00 "Has CM obtained one or more CSEs at grades 2 to 5?". ************** AS AND A-LEVELS *****************************. recode edasl1 (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into aslhi00 . recode edasl2 (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into asllo00 . * Hi is 9 and 7. variable labels aslhi00 "Has CM obtained one or more AS levels at grades A to C?". variable labels asllo00 "Has CM obtained one or more AS levels at grades D to G?". compute Naslev00=edasl1 . recode Naslev00 (missing=0) (98=1) (99=0) . variable labels Naslev00 "Number of AS level passes at grade A-C". recode edgcasl1 (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into alevhi00 . recode edgcasl2 (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into alevlo00 . compute alev00=alevhi00 + alevlo00 . recode alev00 (0=0) (1,2=1) . variable labels alevhi00 "Has CM obtained one or more A levels at grades A to C?". variable labels alevlo00 "Has CM obtained one or more A levels at grades D to E?". variable labels alev00 "Has CM obtained any A level passes at any grade?". compute t1=edgcasl1 . compute t2=edgcasl2 . recode t1 (missing=0) (98=1) (99=0) . recode t2 (missing=0) (98=1) (99=0) . * Someone with a 98 or 99 only gets 1 A-level. compute Nalev00 = t1+t2 . variable labels Nalev00 "Number of A level passes at any grade". ********************** SCOTTISH SCHOOL QUALIFICATIONS *************. *freq edscot1 edscot2 edscot3 edscot4 edscot5 edscot6 .


* The Scottish qualifications are in edscot1 to edscot6 although edscot5 and edscot6 only have missing values. * This is not obvious from documentation. compute scotlo00=0 . compute scot200=0 . compute scot300=0 . compute scot400=0 . compute scot500=0 . compute scot600=0 . variable labels scotlo00 "Has CM obtained Scottish SCE standard grades 4-5 or equivalent?". variable labels scot200 "Has CM obtained Scottish SCE standard grades 1-3 or equivalent?". variable labels scot300 "Has CM obtained Scottish SUPE-SLC lower or ordinary grade?". variable labels scot400 "Has CM obtained Scottish SCE-SUPE-SLC higher grade or equivalent?". variable labels scot500 "Has CM obtained Scottish Certificate of 6th Year Studies?". variable labels scot600 "Has CM obtained other Scottish school qualification?". * Scottish higher equivalent to A level in IFS definitions. if ( any(1,edscot1,edscot2,edscot3,edscot4,edscot5,edscot6) ) scotlo00=1 . if ( any(2,edscot1,edscot2,edscot3,edscot4,edscot5,edscot6) ) scot200=1 . if ( any(3,edscot1,edscot2,edscot3,edscot4,edscot5,edscot6) ) scot300=1 . if ( any(4,edscot1,edscot2,edscot3,edscot4,edscot5,edscot6) ) scot400=1 . if ( any(5,edscot1,edscot2,edscot3,edscot4,edscot5,edscot6) ) scot500=1 . if ( any(6,edscot1,edscot2,edscot3,edscot4,edscot5,edscot6) ) scot600=1 . ************** DEGREES ************************. * Is 9 the missing value for numdeg. recode numdeg (missing=0) (0=0) (9=0) (1 thru hi=1) into deg00 . * Hi is 4. variable labels deg00 "Has CM obtained a degree?". recode numothdg (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into odeg00 . * hi = 6. variable labels odeg00 "Other degree level quals such as graduate membership of professional institutes". compute ALTodeg=0 . if (any(9,edqtyp14,edqtyp15,edqtyp16,edqtyp17,edqtyp18,edqtyp19,edqtyp20,edqtyp21, edqtyp22,edqtyp23,edqtyp24,edqtyp25,edqtyp26) ) ALTodeg=1 . *crosstabs ALTodeg by odeg00 . *crosstabs numothdg by odeg00 . recode numhghdg (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into hdeg00 . *hi = 3, no 98 or 99. variable labels hdeg00 "Has CM obtained a higher degree?". **************** NURSING, DIPLOMAS, PGCE **************. *There are 3 values for edqtyp21, all the values for edqtyp22 to edqtyp26 are missing. compute dip00=0 . if (any(7,edqtyp14,edqtyp15,edqtyp16,edqtyp17,edqtyp18,edqtyp19,edqtyp20,edqtyp21, edqtyp22,edqtyp23,edqtyp24,edqtyp25,edqtyp26) or (numotht > 0 & numotht < 15)) dip00=1 .


variable labels dip00 "Diploma-Certificate-Teacher Training Qual (not pgce) in Higher Ed". * 357 have 1 teacher training qualification. 1 person has 12. recode numparam (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into nurse00 . *hi=15 No 98 or 99. variable labels nurse00 "Has CM obtained any nursing or paramedic qualifications?". compute pgce00=0 . If (any(12,edqtyp14,edqtyp15,edqtyp16,edqtyp17,edqtyp18,edqtyp19, edqtyp20,edqtyp21,edqtyp22,edqtyp23,edqtyp24,edqtyp25,edqtyp26)) pgce00=1 variable labels pgce00 "Has CM obtained a PGCE or other postgraduate certificate in HE?". execute . *************************************************************************************************************. **************************** FIRST BLOCK FOR RUNNING PROGRAM ******************* . *************************************************************************************************************. ********* ONC or OND AND HNC or HND ********************. recode voconc (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into oncond00 . * hi=7 . variable labels oncond00 "Does CM have ONC or OND?". recode vochnc (missing=0) (0=0) (9=0) (98=1) (99=0) (1 thru hi=1) into hnchnd00 . * hi=4. One value of 9. variable labels hnchnd00 "Does CM have HNC or HND?". ***** BTECS *****. compute btec100=0 . compute btec200=0 . variable labels btec100 "Does CM have a BTEC etc First-General Certificate?". variable labels btec200 "Does CM have a BTEC etc First-General Diploma?". compute btec300=0 . compute btec400=0 . variable labels btec300 "Does CM have a BTEC etc National Certificate Diploma?". variable labels btec400 "Does CM have a BTEC etc Higher Certificate Diploma?". compute btec500=0 . variable labels btec500 "Does CM have other BTEC qualification?". if (any(1,bteclev,bteclev2,bteclev3,bteclev4,bteclev5,bteclev6,bteclev7,bteclev8,bteclev9, btecle10,btecle11,btecle12,btecle13,btecle14,btecle15)) btec100=1 . if (any(2,bteclev,bteclev2,bteclev3,bteclev4,bteclev5,bteclev6,bteclev7,bteclev8,bteclev9, btecle10,btecle11,btecle12,btecle13,btecle14,btecle15)) btec200=1 . if (any(3,bteclev,bteclev2,bteclev3,bteclev4,bteclev5,bteclev6,bteclev7,bteclev8,bteclev9, btecle10,btecle11,btecle12,btecle13,btecle14,btecle15)) btec300=1 . if (any(4,bteclev,bteclev2,bteclev3,bteclev4,bteclev5,bteclev6,bteclev7,bteclev8,bteclev9, btecle10,btecle11,btecle12,btecle13,btecle14,btecle15)) btec400=1 .


if (any(5,bteclev,bteclev2,bteclev3,bteclev4,bteclev5,bteclev6,bteclev7,bteclev8,bteclev9, btecle10,btecle11,btecle12,btecle13,btecle14,btecle15)) btec500=1 . * Only one observation in each of bteclev9 to bteclev15. All have first or general certificate. * Some values equal to 9 eg 89 observations with 9 for bteclev. ***** CITY AND GUILDS *****. compute city100=0 . compute city200=0 . compute city300=0 . compute city400=0 . compute city500=0 . variable labels city100 "Does CM have a City and Guilds Part 1 " . variable labels city200 "Does CM have a City and Guilds Part2 or Craft or Intermediate" . variable labels city300 "Does CM have a City and Guilds Part3 or Final or Advanced Craft" . variable labels city400 "Does CM have a City and Guilds Part4 or Career Extension or Full technological" . variable labels city500 "Does CM have a City and Guilds Other C&G qualification" . If (any(1,citylev,citylev2,citylev3,citylev4,citylev5,citylev6,citylev7,citylev8,citylev9,cityle10,cityle11, cityle12,cityle13,cityle14,cityle15)) city100=1 . If ( any(2,citylev,citylev2,citylev3,citylev4,citylev5,citylev6,citylev7,citylev8,citylev9,cityle10,cityle11, cityle12,cityle13,cityle14,cityle15)) city200=1 . If any(3,citylev,citylev2,citylev3,citylev4,citylev5,citylev6,citylev7,citylev8,citylev9,cityle10,cityle11, cityle12,cityle13,cityle14,cityle15) city300=1 . If any(4,citylev,citylev2,citylev3,citylev4,citylev5,citylev6,citylev7,citylev8,citylev9,cityle10,cityle11, cityle12,cityle13,cityle14,cityle15) city400=1 . If any(5,citylev,citylev2,citylev3,citylev4,citylev5,citylev6,citylev7,citylev8,citylev9,cityle10,cityle11, cityle12,cityle13,cityle14,cityle15) city500=1 . **** RSA ****************. compute rsa100=0 . compute rsa200=0 . compute rsa300=0 . compute rsa400=0 . compute rsa500=0 . variable labels rsa100 "Does CM have RSA certificate?". variable labels rsa200 "Does CM have RSA First Diploma?". variable labels rsa300 "Does CM have RSA Advanced Diploma or Certificate?". variable labels rsa400 "Does CM have RSA Higher Diploma" . variable labels rsa500 "Does CM have other RSA qualification?". If (any(1,rsalev,rsalev2,rsalev3,rsalev4,rsalev5,rsalev6,rsalev7,rsalev8,rsalev9,rsalev10, rsalev11,rsalev12)) rsa100=1 . If (any(2,rsalev,rsalev2,rsalev3,rsalev4,rsalev5,rsalev6,rsalev7,rsalev8,rsalev9,rsalev10, rsalev11,rsalev12)) rsa200=1 . If (any(3,rsalev,rsalev2,rsalev3,rsalev4,rsalev5,rsalev6,rsalev7,rsalev8,rsalev9,rsalev10,


rsalev11,rsalev12)) rsa300=1 . If (any(4,rsalev,rsalev2,rsalev3,rsalev4,rsalev5,rsalev6,rsalev7,rsalev8,rsalev9,rsalev10, rsalev11,rsalev12)) rsa400=1 . If (any(5,rsalev,rsalev2,rsalev3,rsalev4,rsalev5,rsalev6,rsalev7,rsalev8,rsalev9,rsalev10, rsalev11,rsalev12)) rsa500=1 . * rsalev10 to rsalev12 only have 1 observation each. All values are for first diplomas (2). **** PITMANS *******. * Pitmans are only in the 2000 data. compute pit100=0 . compute pit200=0 . compute pit300=0 . compute pit400=0 . variable labels pit100 "Does CM have a Pitmans level 1?". variable labels pit200 "Does CM have a Pitmans level 2?". variable labels pit300 "Does CM have a Pitmans level 3?". variable labels pit400 "Does CM have other Pitmans qualification?". If (any(1,pitlev,pitlev2,pitlev3,pitlev4,pitlev5,pitlev6,pitlev7)) pit100=1 . If (any(2,pitlev,pitlev2,pitlev3,pitlev4,pitlev5,pitlev6,pitlev7)) pit200=1 . If (any(3,pitlev,pitlev2,pitlev3,pitlev4,pitlev5,pitlev6,pitlev7)) pit300=1 . If (any(4,pitlev,pitlev2,pitlev3,pitlev4,pitlev5,pitlev6,pitlev7)) pit400=1 . * pitlev7 has one observation with a value of level 2. ******** NVQ *************. *nvqs (current sweep only) . compute nvq100=0 . compute nvq200=0 . compute nvq300=0 . compute nvq400=0 . compute nvq500=0 . compute nvq600=0 . compute nvq700=0 . compute nvq800=0 . variable labels nvq100 "Does CM have an NVQ at Level 1?". variable labels nvq200 "Does CM have an NVQ at Level 2?". variable labels nvq300 "Does CM have an NVQ at Level 3?". variable labels nvq400 "Does CM have an NVQ at Level 4?". variable labels nvq500 "Does CM have an NVQ at Level 5?". variable labels nvq600 "Does CM have an NVQ at Level 6?". variable labels nvq700 "Does CM have Trusts towards NVQ-SVQ?". variable labels nvq800 "Does CM have other NVQ?". If (any(1,nvqlev,nvqlev2,nvqlev3,nvqlev4,nvqlev5,nvqlev6,nvqlev7,nvqlev8,nvqlev9,nvqlev10)) nvq100=1 . If (any(2,nvqlev,nvqlev2,nvqlev3,nvqlev4,nvqlev5,nvqlev6,nvqlev7,nvqlev8,nvqlev9,nvqlev10)) nvq200=1 . If (any(3,nvqlev,nvqlev2,nvqlev3,nvqlev4,nvqlev5,nvqlev6,nvqlev7,nvqlev8,nvqlev9,nvqlev10)) nvq300=1 . If (any(4,nvqlev,nvqlev2,nvqlev3,nvqlev4,nvqlev5,nvqlev6,nvqlev7,nvqlev8,nvqlev9,nvqlev10)) nvq400=1 .


If (any(5,nvqlev,nvqlev2,nvqlev3,nvqlev4,nvqlev5,nvqlev6,nvqlev7,nvqlev8,nvqlev9,nvqlev10)) nvq500=1 . If (any(6,nvqlev,nvqlev2,nvqlev3,nvqlev4,nvqlev5,nvqlev6,nvqlev7,nvqlev8,nvqlev9,nvqlev10)) nvq600=1 . If (any(7,nvqlev,nvqlev2,nvqlev3,nvqlev4,nvqlev5,nvqlev6,nvqlev7,nvqlev8,nvqlev9,nvqlev10)) nvq700=1 . If (any(8,nvqlev,nvqlev2,nvqlev3,nvqlev4,nvqlev5,nvqlev6,nvqlev7,nvqlev8,nvqlev9,nvqlev10)) nvq800=1 . *nvqlev7 to nvqlev10 have no values. ****** GNVQ *****************. * gnvqs (current sweep only) compute gnvq100=0 . compute gnvq200=0 . compute gnvq300=0 . compute gnvq400=0 . variable labels gnvq100 "Does CM have GNVQ Foundation?". variable labels gnvq200 "Does CM have GNVQ Intermediate?". variable labels gnvq300 "Does CM have GNVQ Advanced?". variable labels gnvq400 "Does CM have other GNVQ qualification?". If (any(1,gnvlev,gnvlev2,gnvlev3,gnvlev4,gnvlev5)) gnvq100=1 . If (any(2,gnvlev,gnvlev2,gnvlev3,gnvlev4,gnvlev5)) gnvq200=1 . If (any(3,gnvlev,gnvlev2,gnvlev3,gnvlev4,gnvlev5)) gnvq300=1 . If (any(4,gnvlev,gnvlev2,gnvlev3,gnvlev4,gnvlev5)) gnvq400=1. *gnvlev4 and gnvlev5 have no values. *** OTHER QUALIFICATIONS *************. recode vocappr (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into appren00 . * hi is 7. variable labels appren00 "Has CM obtained any recognised trade apprentice qualifications?". recode vochgv (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into hgv00 . * hi is 3. variable labels hgv00 "Has CM obtained at least one hgv qualification?". recode vocoth (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into othvoc00 . * hi is 15. variable labels othvoc00 "Has CM obtained at least one other vocational qualification?". ********************************************************************************************. ********************** MISSING VALUE ANALYSIS **********************************. ********************************************************************************************. * edquals2, vocqual 1 ='Yes' 2='No' . * We argue that if initial feeder questions (edquals2, vocqual) are answered 'properly' then we know whether the individual has a qualification or not. If the answer is missing or 'do not know' then we do not know. The code below sets all the educational qualifications to missing if edquals is not answered Yes or No or NA and all the vocational qualifications to missing if vocqual is not answered Yes or No or 'Not Applicable'.


recode edquals2 (2=0) (1=0) (8=1) (9=0) (missing=2) into acamiss. variable label acamiss 'Dummy for missing value for do you have an educational qualification'. value label acamiss 0 'not missing' 1 '8 (do not know)' 2 'missing'. recode vocqual (1=0) (2=0) (8=1) (9=0) (missing=2) into vocmiss. * vocmiss adjusted for nurse00 (nurse00 appears as an educational qualification) below. variable label vocmiss 'Dummy for missing value for do you have a vocational qualification'. value label vocmiss 0 'not missing' 1 '8 (do not know)' 2 'missing'. do if (vocmiss eq 0 and acamiss eq 0) . compute allmiss =0 . else if (vocmiss eq 1 and acamiss eq 0) . compute allmiss =1 . else if (vocmiss eq 0 and acamiss eq 1) . compute allmiss =2 . else if (vocmiss eq 1 and acamiss eq 1) . compute allmiss =3 . else if (vocmiss eq 2 and acamiss eq 2) . compute allmiss =4 . end if . variable label allmiss 'Missing values, Do you have any of these qualifications?'. value label allmiss 0 'not missing' 1 '8 vocat quals, not missing for educ quals' 2 ' vocat quals not missing, 8 for educ quals' 3 '8 for both vocat and educ quals' 4 'missing for both vocational & educational quals'. ****** Set missing values **************. do if (acamiss gt 0) . compute gcsehi00 =-9 . compute gcselo00 =-9 . compute olevhi00 = -9 . compute olevlo00 = -9 . compute csehi00 = -9 . compute cselo00 = -9 . compute aslhi00 = -9 . compute asllo00 = -9. compute naslev00 = -9. compute alevhi00 = -9. compute alevlo00 = -9. compute alev00 = -9. compute nalev00 = -9. compute scotlo00 = -9. compute scot200 = -9. compute scot300 = -9. compute scot400 = -9. compute scot500 = -9. compute scot600 = -9. compute dip00 = -9. compute deg00 = -9. compute odeg00 = -9. compute dip00 = -9. compute hdeg00 = -9. compute nurse00 = -9. compute pgce00 = -9 . end if .


if (nurse00 eq 1) vocmiss eq 0 . * nurse is a vocational qualification under some typologies eg nvq but educational here. do if (vocmiss gt 0) . compute oncond00 =-9 . compute hnchnd00 =-9 . compute btec100=-9 . compute btec200=-9 . compute btec300=-9 . compute btec400=-9 . compute btec500=-9 . compute city100 =-9 . compute city200 =-9 . compute city300 =-9 . compute city400 =-9 . compute city500 = -9 . compute rsa100 = -9 . compute rsa200 =-9 . compute rsa300 =-9 . compute rsa400 =-9 . compute rsa500 =-9 . compute pit100 =-9 . compute pit200 =-9 . compute pit300 =-9 . compute pit400 =-9 . compute nvq100 =-9 . compute nvq200 =-9 . compute nvq300 =-9 . compute nvq400 =-9 . compute nvq500 =-9 . compute nvq600 =-9 . compute nvq700 =-9 . compute nvq800 =-9 . compute gnvq100 =-9 . compute gnvq200 =-9 . compute gnvq300 =-9 . compute gnvq400 =-9 . compute appren00=-9 . compute hgv00 =-9 . compute othvoc00 =-9 . end if . missing value gcsehi00 to pgce00 oncond00 to othvoc00 (-9) . compute proxy = 0 . if (edquals gt 0) proxy =1 . * if (vocquals gt 0) proxy1 =1 not needed as proxy = proxy1. variable label proxy 'Was interview completed by a proxy'. * Acamiss=vocmiss=1 when proxy=1 so proxy 'accounts' for 26 of the missing values. compute acamiss1=0 . compute vocmiss1=0 . if (missing(edqtyp14) eq 1 and missing(edqtyp15) eq 1 and missing(edqtyp16) eq 1 and missing(edqtyp17) eq 1 and missing(edqtyp18) eq 1 and missing(edqtyp19) eq 1 and missing(edqtyp20) eq 1 and missing(edqtyp21) eq 1 and missing(edqtyp22) eq 1 and


missing(edqtyp23) eq 1 and missing(edqtyp24) eq 1 and missing(edqtyp25) eq 1 and missing(edqtyp26) eq 1 ) acamiss1=1 . if (missing(voctyp12) eq 1 and missing(voctyp13) eq 1 and missing(voctyp14) eq 1 and missing(voctyp15) eq 1 and missing(voctyp16) eq 1 and missing(voctyp17) eq 1 and missing(voctyp18) eq 1 and missing(voctyp19) eq 1 and missing(voctyp20) eq 1 and missing(voctyp21) eq 1 and missing(voctyp22) eq 1 ) vocmiss1=1 . ********** GOVERNMENT TRAINING SCHEMES **************. recode numyts (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into yts00 . * hi is 15. variable labels yts00 "Has CM done a YTS course?". recode numgov (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into gov00 . * hi is 15. variable labels gov00 "Has CM done another gov course (including New Deal)?". recode numap (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into modapp00 . * hi is 3. variable labels modapp00 "Has CM done a modern apprenticeship?". execute . ******************************************************************************************************. ********************** SECOND BLOCK WHEN RUNNING PROGRAM **********************. ******************************************************************************************************. ************* OTHER COURSES ******************************. recode numfqual (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into nfail00. *Hi =11. variable labels nfail00 "Number of courses started where CM did not obtain the qualification". recode numac (missing=0) (0=0) (98=1) (99=0) (1 thru hi=1) into access00 . * hi is 10. variable labels access00 "Has CM done an access course?". compute workc00=numwrktr . recode workc00 (missing=0) (0=0) (98=1) (99=0) . * max is 30. variable labels workc00 "Number of other work related courses". *************** NON WORK RELATED COURSES ****************. compute leis00=numleis . recode leis00 (missing=0) (0=0) (98=1) (99=0) . * max is 15. variable labels leis00 "Number of courses for interest or leisure".


**** READING, WRITING AND MATHS COURSES ********** . recode numread (missing=0) (0=0) (98=1) (99=0) (1=1) (2=2) (3 thru hi=3) into read00 . * max is 11. variable labels read00 "Number of courses to improve reading". recode numwrite (missing=0) (0=0) (98=1) (99=0) (1=1) (2=2) (3 thru hi=3) into write00 . * max is 11. variable labels write00 "Number of courses to improve writing". * creative writing ?. recode nummaths (missing=0) (0=0) (98=1) (99=0) (1=1) (2=2) (3 thru hi=3) into maths00 . * max is 15. variable labels maths00 "Number of courses to improve maths". value labels read00 write00 maths00 3 "3 or more" . *** THERE IS A LOT ABOUT CURRENT COURSES THAT WE HAVE NOT USED ***********. *******************************************************************************************************. ********************* DUMMIES FOR GROUPS OF QUALIFICATIONS *********************. *******************************************************************************************************. ************* HAS CM OBTAINED A PARTICULAR ACADEMIC QUALIFICATION? ************. do if (acamiss gt 0) . compute Aca000=-9 . compute Aca100=-9 . compute Aca200=-9 . compute Aca300=-9 . compute Aca400=-9 . compute Aca500=-9 . compute Aca600=-9 . compute Aca700=-9 . compute Aca800=-9 . else . compute Aca000=1 . compute Aca100=0 . compute Aca200=0 . compute Aca300=0 . compute Aca400=0 . compute Aca500=0 . compute Aca600=0 . compute Aca700=0 . compute Aca800=0 . end if . missing values Aca000 to Aca800 (-9) . * AcaX00 dummy variables variables showing whether CM has one of a group of qualifications. variable labels Aca000 "1 if CM has no academic qualifications". variable labels Aca100 "1 if CM has bad O-levels ". variable labels Aca200 "1 if CM has CSE 2-5, Other Scots ". variable labels Aca300 "1 if CM has good O-levels ". variable labels Aca400 "1 if CM has 1 A-level or more than 1 AS at grade A-C ". variable labels Aca500 "1 if CM has 2 or more A-levels".


variable labels Aca600 "1 if CM has diploma ". variable labels Aca700 "1 if CM has degree, PGCE, other degree level ". variable labels Aca800 "1 if CM has higher degree". value labels aca000 0 "At least one academic qualification" 1 "No academic qualifications". * Bad O-levels GCSEs D to E, GCEs D to E, Scottish SCE standard grades 4-5' . * Good O-levels CSE grade 1, GCSE grades A-C, GCE grades A-C, Scots standard grades 1-3, Scots lower or ordinary. * 2 or more A-LEVELS 2 or more A-levels, Scottish higher grade or Certificate of 6th Year Studies . do if (hdeg00 =1) . compute Aca800=1 . compute Aca000=0 . end if . do if (any(1,deg00,odeg00,pgce00)) . compute Aca700=1 . compute Aca000=0 . end if . do if (dip00 eq 1) . compute Aca600=1 . compute Aca000=0 . end if . do if (Nalev00 gt 1 or scot400 eq 1 or scot500 eq 1) . compute Aca500=1 . compute Aca000=0 . end if . do if (Nalev00 eq 1 or Naslev00 gt 1) . compute Aca400=1 . compute Aca000=0 . end if . do if (any(1,gcsehi00,olevhi00,csehi00,scot200,scot300)) . compute Aca300=1. compute Aca000=0 . end if . do if (cselo00 eq 1 or scot600 eq 1) . compute Aca200=1 . compute Aca000=0 . end if . do if (any(1,gcselo00,olevlo00,scotlo00)) . compute Aca100=1 . compute Aca000=0 . end if . ************* HAS CM OBTAINED A PARTICULAR VOCATIONAL QUALIFICATION? ************. do if (Vocmiss gt 0) . compute Voc000=-9 . compute Voc100=-9 .


compute Voc200=-9 . compute Voc300=-9 . compute Voc400=-9 . compute Voc500=-9 . else . compute Voc000=1 . compute Voc100=0 . compute Voc200=0 . compute Voc300=0 . compute Voc400=0 . compute Voc500=0 . end if . missing values Voc000 to Voc500 (-9) . * VocX00 dummy variables variables showing whether CM has one of a group of qualifications. variable labels Voc000 "1 if CM has no vocational qualifications". variable labels Voc100 "1 if CM has vocational qualification equivalent to NVQ1". variable labels Voc200 "1 if CM has vocational qualification equivalent to NVQ2". variable labels Voc300 "1 if CM has vocational qualification equivalent to NVQ3". variable labels Voc400 "1 if CM has vocational qualification equivalent to NVQ4". variable labels Voc500 "1 if CM has vocational qualification equivalent to NVQ5". value labels voc000 0 "At least one vocational qualification" 1 "No vocational qualifications". * 100 'C&G Opo & Other, RSA 1, HGV, Other tech or bus qualification' 200 'C&G Craft-Inter-Ord-Part1-Cant say, RSA2, Insig,JIB NJC etc' 300 'C&G Advanced-Final-Part II or III, ONC-OND, RSA 3 ' 400 'Nursing, BTEC HNC-HND, HNC-HND, C&G Full Tech (FTC)' 500 'NVQ awarded at level 5'. do if (any(1,nvq500,nvq600)) . compute Voc500=1 . compute Voc000=0 . end if . do if (any(1,nurse00,hnchnd00,btec400,rsa400,nvq400)) . compute Voc400=1 . compute Voc000=0 . end if . do if (any(1,city200,city300,city400,oncond00,btec300,rsa300,pit300,nvq300,gnvq300)) . compute Voc300=1 . compute Voc000=0 . end if . do if (any(1,city100,btec200,rsa200, pit200,appren00,nvq200,gnvq200)) . compute Voc200=1 . compute Voc000=0 . end if . do if (any(1,city500,btec100,rsa100,pit100,nvq100,gnvq100,hgv00,othvoc00,rsa500, btec500,pit400,nvq700,nvq800,gnvq400)) . compute Voc100=1 . compute Voc000=0 . end if .


********************************************************************************************. ********************* HIGHEST QUALIFICATIONS ****************************. ********************************************************************************************. * This code computes the highest academic qualification, the highest vocational qualification and the highest NVQ level recorded in the 2000 survey. * Previous published work has used slightly different definitions of these variables so you may need to adapt the code below. * Check bad O-levels, many descriptions give them NVQ level of 0 . * All vocational stuff needs checking especially pitmans btec rsa . ********************* HIGHEST ACADEMIC QUALIFICATION ***********************. do if (acamiss gt 0) . compute Hiaca00=-9 . else if (Aca800=1) . compute Hiaca00=8 . else if (Aca700=1) . compute Hiaca00=7 . else if (Aca600=1) . compute Hiaca00=6 . else if (Aca500=1) . compute Hiaca00=5 . else if (Aca400=1) . compute Hiaca00=4 . else if (Aca300=1) . compute Hiaca00=3 . else if (Aca200=1) . compute Hiaca00=2 . else if (Aca100=1) . compute Hiaca00=1 . else if (acamiss eq 0) . compute Hiaca00=0 . end if . variable label Hiaca00 'Highest academic qualification recorded in 2000 survey'. value labels Hiaca00 -9 Missing 0 'None ' 1 ' Bad O-levels' 2 'CSE 2-5, other Scottish school qualification' 3 'Good O levels' 4 '1 A level or more than 1 AS level at grade A-C' 5 '2 or more A-levels' 6 'Diploma' 7 'Degree, PGCE, other degree level'


8 'Higher degree ' . * Notice diploma=degree . recode Hiaca00 (-9=-9) (0=0) (1=1) (2=1)(3=2)(4=2)(5=3)(6=4)(7=4)(8=5) into nvqaca00 . variable label nvqaca00 'Highest NVQ level from an academic qualification in 2000 survey'. recode Hiaca00 (-9=-9) (0=0) (1=1) (2=1)(3=2)(4=2)(5=3)(6=4)(7=5)(8=6) into ghmaca00 . variable label ghmaca00 'Highest academic qualification in 2000 survey ghm measure see also IFS'. value labels ghmaca00 0 'None' 1 'Bad O levels, CSE 2-5' 2 'Good O-levels, 1 A-level' 3 '2 or more A-levels' 4 'Sub-degree' 5 'Degree' 6 'Higher Degree' . ********************* HIGHEST VOCATIONAL QUALIFICATION ***********************. do if (vocmiss gt 0) . compute HiVoc00=-9 . else if (Voc500 eq 1) . compute HiVoc00=5 . else if (Voc400 eq 1) . compute HiVoc00=4 . else if (Voc300 eq 1) . compute HiVoc00=3 . else if (Voc200 eq 1) . compute HiVoc00=2 . else if (Voc100 eq 1) . compute HiVoc00=1 . else if (vocmiss eq 0) . compute HiVoc00=0 . end if . variable label HiVoc00 'Highest vocational qualification recorded in 2000 survey'. * HiVoc00 is also the highest NVQ level from a vocational qualification in 2000 survey. value labels HiVoc00 -9 Missing 0 'None ' 1 'C&G Opo & Other, RSA 1, HGV, Other tech or bus qualification' 2 'C&G Craft-Inter-Ord-Part1-Cant say, RSA2, Insig,JIB NJC etc' 3 'C&G Advanced-Final-Part II or III, ONC-OND, RSA 3 ' 4 'Nursing, BTEC HNC-HND, HNC-HND, C&G Full Tech (FTC)' 5 'NVQ awarded at level 5'. *********************** Highest NVQ whatever type ********************************. compute HiNVQ00=max(nvqaca00,HiVoc00) .


variable label HiNVQ00 'Highest NVQ level (whether academic or vocational)' . missing value Hiaca00 nvqaca00 ghmaca00 HiVoc00 HiNVQ00 (-9) . *********************** OTHER VARIABLES *******************************************. do if (sample eq 1) . compute group = 1. else if (sample eq 2 and dmpart eq 1) . compute group = 2. else if (sample eq 2 and dmpart eq 2) . compute group = 3. end if . variable label group "3 categories - BCS70 and NCDS by whether present in 1991 survey" . value label group 1 "BCS70" 2 "NCDS present in 1991" 3 "NCDS not present in 1991" . recode dmsex (1=0) (2=1) into female . value labels female 0 "male" 1 "female" . **********************************************************************************************. ******************************* SAVE DATA **********************************************. **********************************************************************************************. save outfile = 'e:\00\ncds-bcs70\qca data\all2000.sav' / drop ALTodeg t1 t2 vocmiss1 acamiss1 . ******************************************************************************************************. *************** THIRD BLOCK WHEN RUNNING PROGRAM **********************. ******************************************************************************************************. ************************* FINISH *********************************************.


2. Samantha Parsons **JUNE 2000 **MERGING NCDS5 QUALIFICATION DATA WITH 2000 DATA. **CHANGED 2000 TO MATCH NCDS33. **simple way of recoding highest academic/vocational qualifications - to match NCDS at 33 hqual vars. **ghmaca00 - levels of academic qualifications: ranges from 0 to 6: with 5 = degree and 6 = higher degree. . **hivoc00 - levels of vocational NVQ qualifications: ranges from 0 to 5: with 5 = NVQ level 5. **variable HINVQ00 combines the information from the two variables above with a range of 0 to 5, BUT level 5 represents ghmaca00 level 6 and hivoc00 level 5 only. Surely it should be ghmaca00 level 5 AND 6 and hivoc00 level 5?. **made such a variable . compute hq2ka = ghmaca00. if (hivoc00 > ghmaca00) hq2ka = hivoc00. recode hq2ka (5,6=5). ** Highest qualification at NCDS5 (age 33) – NB: Uses NCDS5 variables n501513 to n501541 do rep x=n501441 to n501469/y=n501513 to n501541. if x=37 and range (y,1,36) x=y. if x=37 and missing (y) or y=0 y=37. end rep. compute hqual33=-2. do rep x=n501441 to n501469. if any (x,37) hqual33=0. if (any (x,10,25,1) and hqual33 < 1) hqual33=1. if (any (x,20,19,18,17,13,14,12,11,7,6,4,3,2) and hqual33 < 2) hqual33=2 if (any (x,23,21,15,09,08,05) and hqual33 < 3) hqual33=3. if (any (x,30,29,28,27,26,24,22,16) and hqual33 < 4) hqual33=4. if (range (x,31,33) and hqual33 < 5) hqual33=5. end rep. var labels hqual33 "Highest qual gained at age 33 – based on NCDS5 variables ". value labels hqual33 0 ‘No qualification’ 1 ‘CSE 2-5/equiv NVQ1’ 2 ‘O Level/equiv NVQ2’ 3 ‘A Level/equiv NVQ3’ 4 ‘Higher qual NVQ4’ 5 ‘Degree/higher NVQ5/6’ -1 ‘No information’. formats hqual33 f2.0). recode hqual33 (-2=-1) missing values hqual33 (-1). **DIRTY highest qualification variable combining information of NCDS cohort at age 33. compute hq332k = hq2ka. if (hq2ka = 0 and hqual33 >=0) hq332k = hqual33. do if (hq2ka > 0 and hqual33 >= 0). if (hqual33 < hq2ka) hq332k = hq2ka. if (hqual33 = hq2ka) hq332k = hq2ka. if (hqual33 > hq2ka) hq332k = hqual33.


end if. variable labels hq332k 'highest NVQ qualification - academic or vocational - recoding 2000 to match NCDS5'. frequencies variables=hq332k. *********************************************************************************************************************. **some qualifications coded differently at each sweep, so made individual qualification vars at age 33 to re-create various highest qualification variables at 33 to match 2000. *********************************************************************************************************************. do if (n501441 >= 0). compute noquals = 0. compute cse2_5 = 0. compute cse1 = 0. compute oleva_c = 0. compute gcsea_c = 0. compute alev = 0. compute scotoa_c = 0. compute scots1_3 = 0. compute scothg = 0. compute scot6th = 0. compute rsa1 = 0. compute rsa2 = 0. compute rsa3 = 0. compute cg_op = 0. compute cg_cio1 = 0. compute cg_afp23 = 0. compute cg_ft = 0. compute cg_oth = 0. compute cg_dk = 0. compute cg_ia = 0. compute jibnjcct = 0. compute onchnd = 0. compute hnchnd = 0. compute bt_ngcd = 0. compute bt_hhncd = 0. compute othtbq = 0. compute fprofq = 0. compute pprofq = 0. compute nurse = 0. compute cert = 0. compute hdippgce = 0. compute degree = 0. compute pgraddip = 0. compute hdegree = 0. do repeat x = n501441 to n501469. if (x = 37) noquals = 1. if (x = 1) cse2_5 = 1. if (x = 2) cse1 = 1. if (x = 3) oleva_c = 1. if (x = 4) gcsea_c = 1. if (x = 5) alev = 1. if (x = 6) scotoa_c = 1. if (x = 7) scots1_3 = 1.


if (x = 8) scothg = 1. if (x = 9) scot6th = 1. if (x = 10) rsa1 = 1. if (x = 11) rsa2 = 1. if (x = 12) rsa3 = 1. if (x = 13) cg_op = 1. if (x = 14) cg_cio1 = 1. if (x = 15) cg_afp23 = 1. if (x = 16) cg_ft = 1. if (x = 17) cg_oth = 1. if (x = 18) cg_dk = 1. if (x = 19) cg_ia = 1. if (x = 20) jibnjcct = 1. if (x = 21) onchnd = 1. if (x = 22) hnchnd = 1. if (x = 23) bt_ngcd = 1. if (x = 24) bt_hhncd = 1. if (x = 25) othtbq = 1. if (x = 26) fprofq = 1. if (x = 27) pprofq = 1. if (x = 28) nurse = 1. if (x = 29) cert = 1. if (x = 30) hdippgce = 1. if (x = 31) degree = 1. if (x = 32) pgraddip = 1. if (x = 33) hdegree = 1. end repeat. end if. frequencies variables=noquals cse2_5 cse1 oleva_c gcsea_c alev scotoa_c scots1_3 scothg scot6th rsa1 rsa2 rsa3 cg_op cg_cio1 cg_afp23 cg_ft cg_oth cg_dk cg_ia jibnjcct onchnd hnchnd bt_ngcd bt_hhncd othtbq fprofq pprofq nurse cert hdippgce degree pgraddip hdegree . *********************************************************************************************************************************. **making highest qualification variables to best match 2000 data categories. **a levels are treated as equivalent to 2 plus alevels in 2000 as th enumber of a levels not collected in NCDS5. *********************************************************************************************************************************. *******************************************************************************. **highest academic qualification to match 8 category HIACA00. **a levels in ncds5 are treated as equivalent to 2 plus alevels in 2000. *******************************************************************************. compute hacad33b = 0. if (any(1,cse2_5)) hacad33b = 2. if (any(1,cse1,oleva_c,gcsea_c,scotoa_c,scots1_3)) hacad33b = 3. if (any(1,alev,scothg,scot6th)) hacad33b = 5. if (any(1,cert,hdippgce)) hacad33b = 6. if (any(1,degree,pgraddip)) hacad33b = 7. if (hdegree = 1) hacad33b = 8. if (missing(noquals)) hacad33b = -1. missing values hacad33b (-1).


variable labels hacad33b 'ncds33: highest academic qualification to match HIACA00'. value labels hacad33b 0'no quals' 2'cse2-5' 3'good o levels' 5'a levels' 6'diploma' 7'degree PGCE' 8'higher degree'. frequencies variables=hacad33b. ********************************************************************************************. **highest academic qualification variable - NVQ levels - to match NVQACA00. ********************************************************************************************. recode hacad33b (0=0) (2=1) (3=2) (5=3) (6,7=4) (8=5) into hacnvq33. variable labels hacnvq33 'ncds33: highest NVQ level qual from academic qualification to match NVQACA00'. frequencies variables=hacnvq33. ********************************************************************************************. **highest academic qualification variable - ghm measure to match GHMACA00. ********************************************************************************************. recode hacad33b (0=0) (2=1) (3=2) (5=3) (6=4) (7=5) (8=6) into hacghm33. variable labels hacghm33 'ncds33: highest academic qualification - ghm measure to match GHMACA00'. value labels hacghm33 0'no quals' 1'cse2-5' 2'o levels' 3'a levels' 4'sub degree' 5'degree' 6'higher degree'. frequencies variables=hacghm33. **************************************************************************************************************. **highest vocational qualification at 33 - including nursing - to NVQ levels - to match HIVOC00'. **************************************************************************************************************. compute hvoc33 = 0. if (any(1,rsa1,othtbq,cg_oth,cg_dk)) hvoc33 = 1. if (any(1,rsa2,cg_op,cg_cio1,cg_ia,jibnjcct)) hvoc33 = 2. if (any(1,rsa3,cg_afp23,onchnd,bt_ngcd)) hvoc33 = 3. if (any(1,cg_ft,hnchnd,bt_hhncd,fprofq,pprofq,nurse)) hvoc33 = 4. if (missing(noquals)) hvoc33 = -1. missing values hvoc33 (-1). variable labels hvoc33 'ncds33: highest vocational qualification - to match HIVOC00'. frequencies variables=hvoc33. ***********************************************************************************************. **highest academic or vocational qualification - to NVQ levels - to match HINVQ00. ************************************************************************************************. compute hq33a = 0. if (any(1,cse2_5,rsa1,othtbq,cg_oth,cg_dk)) hq33a = 1. if (any(1,cse1,oleva_c,olevd_e,scotoa_c,scots1_3,rsa2,cg_op,cg_cio1,cg_ia,jibnjcct)) hq33a = 2. if (any(1,alev,scothg,scot6th,rsa3,cg_afp23,onchnd,bt_ngcd)) hq33a = 3.


if (any(1,cg_ft,hnchnd,bt_hhncd,fprofq,pprofq,nurse,cert,hdippgce,degree,pgraddip)) hq33a = 4. if (hdegree = 1) hq33a = 5. if (missing(noquals)) hq33a = -1. missing values hq33a (-1). variable labels hq33a 'ncds33: highest NVQ qualification - academic or vocational - to match HINVQ00'. frequencies variables=hq33a. *****************************************************************. **COMBINING NCDS33 AND 2000 DATA VARIABLES. *****************************************************************. ****************************************************************************. **combined highest academic qualification variables - 8 category. ****************************************************************************. compute hqaca00 = HIACA00. if (HIACA00 = 0 and hacad33b >=0) hqaca00 = hacad33b. do if (HIACA00 > 0 and hacad33b >= 0). if (hacad33b < HIACA00) hqaca00 = HIACA00. if (hacad33b = HIACA00) hqaca00 = HIACA00. if (hacad33b > HIACA00) hqaca00 = hacad33b. end if. variable labels hqaca00 'highest academic qualification - combining hacad33b and hiaca00'. value labels hqaca00 0'no quals' 1'bad o levels' 2'cse2-5' 3'good o levels' 4'as levels or 1 a level' 5'a levels' 6'diploma' 7'degree PGCE' 8'higher degree'. frequencies variables=hacad33b hiaca00 hqaca00. ****************************************************************************. **combined highest academic qualification variables - NVQ levels. ****************************************************************************. compute hqanvq00 = nvqaca00. if (nvqaca00 = 0 and hacnvq33 >= 0) hqanvq00 = hacnvq33. do if (nvqaca00 > 0 and hacnvq33 >= 0). if (nvqaca00 > hacnvq33) hqanvq00 = nvqaca00. if (nvqaca00 = hacnvq33) hqanvq00 = nvqaca00. if (nvqaca00 < hacnvq33) hqanvq00 = hacnvq33. end if. variable labels hqanvq00 'highest NVQ level qual from academic qualification - combining hacnvq33 and nvqaca00'. frequencies variables=hqanvq00 nvqaca00 hacnvq33 . *******************************************************************************. **combined highest academic qualification variables - ghm measure . *******************************************************************************.


compute hqaghm00 = ghmaca00. if (ghmaca00 = 0 and hacghm33 >= 0) hqaghm00 = hacghm33. do if (ghmaca00 > 0 and hacnvq33 >= 0). if (ghmaca00 > hacghm33) hqaghm00 = ghmaca00. if (ghmaca00 = hacghm33) hqaghm00 = ghmaca00. if (ghmaca00 < hacghm33) hqaghm00 = hacghm33. end if. variable labels hqaghm00 'highest academic qualification - ghm measure - combining hacghm33 and ghmaca00'. value labels hqaghm00 0'no quals' 1'cse2-5' 2'o levels' 3'a levels' 4'sub degree' 5'degree' 6'higher degree'. frequencies variables=hacghm33 ghmaca00 hqaghm00. **************************************. **highest vocational qualification. **************************************. compute hqvoc00 = hivoc00. if (hivoc00 = 0 and hvoc33 >= 0) hqvoc00 = hvoc33. do if (hivoc00 > 0 and hvoc33 >= 0). if (hivoc00 > hvoc33) hqvoc00 = hivoc00. if (hivoc00 = hvoc33) hqvoc00 = hivoc00. if (hivoc00 < hvoc33) hqvoc00 = hvoc33. end if. variable labels hqvoc00 'highest vocational qualification - combining hvoc33 and hivoc00'. frequencies variables=hvoc33 hivoc00 hqvoc00. ***********************************************************************. **highest academic or vocational qualification - to NVQ levels. ***********************************************************************. compute hqnvq00 = hinvq00. if (hinvq00 = 0 and hq33a >= 0) hinvq00 = hq33a. do if (hinvq00 > 0 and hq33a >= 0). if (hinvq00 > hq33a) hqnvq00 = hinvq00. if (hinvq00 = hq33a) hqnvq00 = hinvq00. if (hinvq00 < hq33a) hqnvq00 = hq33a. end if. variable labels hqnvq00 'highest NVQ qualification - academic or vocational - combining hq33a and hinvq00'. frequencies variables=hq33a hinvq00 hqnvq00.


APPENDIX 8: Attitude Scales Documentation relating to the development of attitude scales produced by John Preston of the Centre for Longitudinal Studies is reproduced below:


([email protected])

Attitude Scales for latest sweeps of NCDS and BCS70 Methodology Attitude scales were constructed using the 50 items in the ‘views’ section of the self-completion questionnaire. Each survey was analysed separately, although the factors identified in each were, fortuitously, identical. Principal components analysis was employed using varimax rotation to orthogonal simple structure. The number of factors chosen was based upon those with Eigen values greater than one, followed by the ‘scree’ test to determine that point at which Eigen values leveled off. Factors were examined for interpretability and to ensure that at least three items had loadings on each of them. Cronbach’s alpha was employed as a test of internal consistency of the factors. In all cases, alpha scores were above 0.5 which indicates that it is acceptable to consider the items as forming an identifiable factor. Factor loadings are shown in each table, values less than +/- 0.4 have been suppressed. For each factor, a derived variable has been created which is the mean of the scores for each item within the component. The name of this derived variable is displayed in bold type. Certain items on the factor were inverted prior to the calculation and this is indicated beneath each table. Factors identified 1. Anti-racism TOLERACE

BCS70 NCDS

AR2 Wouldn’t mind if family of different race moved next door

.826 .791

AR4 Wouldn’t mind working with people from other races

.774 .745

AR3 Wouldn’t mind kids going to school with other races

.755 .724

AR5 Would not want another race person as my boss

-.688 -.675

AR1 Mixed race marriage is OK .676

.663

All scale items, with the exception of AR5 were inverted prior to construction of the scale which indicates the level of support for tolerant views related to other ethnic groups.


2. Left-Right LEFTR

BCS70 NCDS

LR5 Ordinary people don’t get a fair share of the nations wealth.

.651 .692

LR6 Government should redistribute income.

.648 .655

LR7 One law for the rich and one for the poor? .612 .641 LR1 Big business benefits owners at the expense of workers.

.572 .605

LR2 Private schools should be abolished .562

.583

LR3 Management get the better of employees .500

.567

The magnitude of the score on LEFTR indicates the level of sympathy with right wing views in terms of redistribution of income, privatization and support for business and management. 3. Support for Authority AUTHORIT

BCS70 NCDS

A4 Give law breakers stiffer sentences .661

.675

A6 Schools teach children to authority .613

.627

A2 Death penalty for some crimes .566

.593

A5 Young people don’t have respect for traditional values

.564 .560

A3 Censorship is needed to uphold morals .481

.463

A1 The law should be obeyed even if it is wrong .442 .441 All items were inverted prior to the construction of a scale of support for authority. 4. Support for traditional marital values PROFAM

BCS70 NCDS

MOR4 Marriage is for life .647

.718

MOR6 Alright for unmarried people to have kids -.629 -.620 MOR3 Couples with kids should not separate .592

.618

MOR1 Divorce is too easy to get these days

.533 .544

MOR5 Women should have the right to an abortion

-.479 -.519

MOR2 Married people happier than unmarried .478 .430 Items MOR4, MOR3, MOR1 and MOR2 were inverted prior to the calculation of this scale which indicates the level of support for marriage and traditional family values.


5. Effect of children on quality of life NOKIDS

BCS70 NCDS

C4 People with no kids are missing out. .756

.731

C1 Unless you have kids, you will be lonely in your old age.

.690 .688

C2 Can have fulfilling life with no kids.

-.661 -.652

C2 was inverted to C2inv prior to the calculation of the scale. NOKIDS indicates the level of agreement with statements opposing the value of children in life experiences. The higher the value of NOKIDS, the greater the level of agreement with these statements. 5. Support for working Mothers MUMWORK

BCS70 NCDS

WM3 Kids benefit if mum has job outside the home

.702 .716

WM4 A mother and family are happier if she goes out to work

.700 .713

WM1 Pre-school kids suffer if mum works

-.607 -.611

WM2 Family life suffers if mum is working full time -.607 -.594 WM5 Dads job is to earn money; mums to stay home

-.408 <-0.4

WM3 and WM4 were inverted prior to the calculation of the scale. MUMWORK indicates agreement with Mothers working. Higher scores indicate that respondents believe that women should work. Note that in the PCA for NCDS, the factor loading for WM5 was of a scalar magnitude less than 0.4. 6. Political cynicism POLITICAL CYNICISM (POLCYN)

BCS70 NCDS

PC2 No difference which political party is in power .769 .769 PC1 No political party would benefit me

.703 .710

PC3 Politicians are in politics for their own benefit .630 .676 Scores on POLCYN indicate agreement with statements regarding the futility and redundancy of the current political system. All scale items were inverted prior to construction of the scale. 7. Technophobia INFORMATION TECHNOLOGY (ANTITECH) BCS70 NCDS IT2 Computers enrich the lives of users .717 .739

IT4 Every family should have a computer .693 .670

IT1 Computers at work are destroying peoples skills

-.563 -.534

IT5 Learning to use a computer is more trouble than it is worth

-.477 -.479

IT1 and IT5 were inverted prior to the creation of the attitude scale. The higher the score on ANTITECH, the greater the degree of opposition to information technology.


8. Environmentalism ENVIRONMENTALISM (ENVIRON)

BCS70 NCDS

E3 The environment vs. economic growth

.722 .717

E2 Preserving the environment is most important. .691 .705 E1 Problems in the environment are not that serious

-.640 -.589

E2 and E3 were inverted to E1inv prior to the construction of the attitude scale. The higher the score on ENVIRON, the greater the degree of individuals support for environmental issues. 9. Support for the work ethic WORK ETHIC (WORKETH)

BCS70 NCDS

WE3 It is important to hang on to any job even if unhappy

.765 .777

WE2 If I didn’t like a job I’d pack it in -.756

-.758

WE1 Any job is better than being unemployed .567 .551 WE1 and WE3 were inverted before construction of the scale which indicates an individuals support for the work ethic. Alpha coefficients SCALE

BCS70 NCDS

Anti-Racism (TOLERACE)

0.8222 0.8000

Left-Right (LEFTR)

0.6792 0.7313

Support for authority (AUTHORIT)

0.6176 0.6470

Support for traditional marital values (PROFAM)

0.6345 0.6620

Effect of children on quality of life (NOKIDS) 0.6232 0.6141 Support for working mothers (MUMHOME) 0.6782 0.6880 Political cynicism (POLCYN)

0.6526 0.6742

Technophobia (ANTITECH)

0.5732 0.5892

Environmentalism (ENVIRON)

0.5070 0.4664

Support for the Work ethic (WORKETH)

0.5469 0.5388


Items not included in scales The following items were not included in the above scales and did not appear to form any clear factors of their own:- C3 Having children interferes with parents freedom L2 Knowing the right people helps more than qualifications in getting a job L4 Effort of getting qualifications is more trouble than it is worth L1 More likely to get better job if do education training IT5 Learning to use a computer is more trouble than it is worth LR4 Take out own health care, stop relying on the NHS John Preston 21/04/2001


SPSS Code MISSING VALUES lr1 lr2 lr3 lr4 lr5 lr6 lr7 ar1 ar2 ar3 ar4 ar5 e1 e2 e3 a1 a2 a3 a4 a5 a6 c1 c2 c3 c4 pc1 pc2 pc3 l1 l2 l3 l4 mor1 mor2 mor3 mor4 mor5 mor6 wm1 wm2 wm3 wm4 wm5 it1 it2 it4 it5 we1 we2 we3 ("8." "9."). COMPUTE tolerace = MEAN((6-ar2),(6-ar4),(6-ar3),ar5,(6-ar1)) . EXECUTE . COMPUTE leftr = MEAN(lr5,lr6,lr7,lr1,lr2,lr3) . EXECUTE . COMPUTE authorit = MEAN((6-a4),(6-a6),(6-a2),(6-a5),(6-a3),(6-a1)) . EXECUTE . COMPUTE profam = MEAN((6-mor4),mor6,(6-mor3),(6-mor1),mor5,(6-mor2)) . EXECUTE . COMPUTE nokids = MEAN(c4,c1,(6-c2)) . EXECUTE . COMPUTE mumwork = MEAN((6-wm3),(6-wm4),wm1,wm2,wm5) . EXECUTE . COMPUTE polcyn = MEAN((6-pc2),(6-pc1),(6-pc3)) . EXECUTE . COMPUTE antitech = MEAN(it2,it4,(6-it1),(6-it5)) . EXECUTE . COMPUTE environ = MEAN((6-e3),(6-e2),e1) . EXECUTE . COMPUTE worketh = MEAN((6-we1),we2,(6-we3)) . EXECUTE . SUMMARIZE /TABLES=ar2 ar4 ar3 ar5 ar1 tolerace /FORMAT=VALIDLIST CASENUM TOTAL LIMIT=50 /TITLE='Case Summaries' /MISSING=VARIABLE /CELLS=COUNT . SUMMARIZE /TABLES=lr5 lr6 lr7 lr1 lr2 lr3 leftr /FORMAT=VALIDLIST CASENUM TOTAL LIMIT=50 /TITLE='Case Summaries' /MISSING=VARIABLE /CELLS=COUNT . SUMMARIZE /TABLES=a4 a6 a2 a5 a3 a1 authorit /FORMAT=VALIDLIST CASENUM TOTAL LIMIT=50 /TITLE='Case Summaries' /MISSING=VARIABLE /CELLS=COUNT . SUMMARIZE /TABLES=mor4 mor6 mor3 mor1 mor5 mor2 profam /FORMAT=VALIDLIST CASENUM TOTAL LIMIT=50 /TITLE='Case Summaries' /MISSING=VARIABLE /CELLS=COUNT . SUMMARIZE /TABLES=c4 c1 c2 nokids /FORMAT=VALIDLIST CASENUM TOTAL LIMIT=50 /TITLE='Case Summaries' /MISSING=VARIABLE /CELLS=COUNT . SUMMARIZE /TABLES=wm3 wm4 wm1 wm2 wm5 mumwork /FORMAT=VALIDLIST CASENUM TOTAL LIMIT=50


/TITLE='Case Summaries' /MISSING=VARIABLE /CELLS=COUNT . SUMMARIZE /TABLES=pc2 pc1 pc3 polcyn /FORMAT=VALIDLIST CASENUM TOTAL LIMIT=50 /TITLE='Case Summaries' /MISSING=VARIABLE /CELLS=COUNT . SUMMARIZE /TABLES=it2 it4 it1 it5 antitech /FORMAT=VALIDLIST CASENUM TOTAL LIMIT=50 /TITLE='Case Summaries' /MISSING=VARIABLE /CELLS=COUNT . SUMMARIZE /TABLES=e3 e2 e1 environ /FORMAT=VALIDLIST CASENUM TOTAL LIMIT=50 /TITLE='Case Summaries' /MISSING=VARIABLE /CELLS=COUNT . SUMMARIZE /TABLES=we3 we2 we1 worketh /FORMAT=VALIDLIST CASENUM TOTAL LIMIT=50 /TITLE='Case Summaries' /MISSING=VARIABLE /CELLS=COUNT .

Documents

Guide to the Combined Dataset - Longitudinal · PDF fileField administration: Elaine Iffland, Theresa Patterson, Glenis Naudo, Jill Penn, Maureen Slater CAPI/CASI programming: